Podcasts about hbm

Announcing the CTP for SpaceX. MahJong Craze gone wild. Goodbye to Alan Greenspan – The Maestro. Have you seen RAM prices? PLUS we are now on Spotify and Amazon Music/Podcasts! Click HERE for Show Notes and Links DHUnplugged is now streaming live - with listener chat. Click on link on the right sidebar. Love the Show? Then how about a Donation? PayPal.Donation.Button({ env:'production', hosted_button_id:'JJJHP2GDEJC7J', image: { src:'https://www.paypalobjects.com/en_US/i/btn/btn_donateCC_LG.gif', alt:'Donate with PayPal button', title:'PayPal - The safer, easier way to pay online!', } }).render('#donate-button'); Follow John C. Dvorak on Twitter Follow Andrew Horowitz on Twitter Warm-Up - Announcing the CTP for SpaceX - MahJong Craze - Goodbye to Alan Greenspan - The Maestro - Have you seen RAM prices? Markets - Economic Collapse Imminent? - Breathe is narrowing again - chips chips chips are the only play - Spacex coming back down to earth? What is that sucking sound? -- Markets getting weird..... 3% down for NASDAQ 100 today - 8% for SMH and 14% for Memory ETF - Just announced - Alphabet (Google) will replace Verizon in DJIA DEDICATION: Alan Greenspan - Died Monday at age 100 Google Enters DJIA - High priced shares - Moves tech to 22% of DJIA from 17% or so - very meaningful move - Every $1 move for Google = $7 move on DJIA - Tech:  S&P 500 (~30%+), Nasdaq (~50%+) Computer Pricing - What as $2,000 a year ago for a nice desktop is not like $4,000 - Dell not holding pricing quotes - and even if they do, back ordered so prices could go up after order - Will IPOs put more money in the pocket of tech companies to buy gear at any price? Endless - SpaceX recently finalized two massive, multibillion-dollar artificial intelligence contracts: a $6.3 billion computing power agreement with Reflection AI and a $60 billion acquisition of the AI coding startup Cursor. - AI Compute Deal with Reflection AI - - - - The Terms: Reflection AI agreed to pay SpaceXAI $150 million per month from July 2026 through the end of 2029. - - -- - - The Infrastructure: The startup will tap into hardware and GB300 chips housed at SpaceX's Colossus 2 data center in Memphis, Tennessee. More SpaceX - SpaceX shares were as high as $220 post IPO. - Sharea ahve been down over the past 3 days. - Most that got in POST IPO probably bought in at about $162-$165 - Newsline: SpaceX shares slipped for a third straight day, shedding hundreds of billions of dollars in market value, after the company said it is selling investment-grade bonds for the first time. - The stock fell 16% Monday to close at $154.60, the lowest level since the company's first day of trading, pushing its three-day loss to 23% and erasing over $600 billion in value over that period. - SpaceX is seeking to raise at least $20 billion from the first bond offering to fund its artificial-intelligence ambitions. Missed Opportunity - Short the Mattress companies he said...... ----- Got squeezed out....Never to return Swing and a Miss Maybe Because this can happen... - Shares of Getty Images Holdings Inc. soared as much as 145% on Monday after it announced a licensing deal with OpenAI. - Getty said that images from its library will appear in the search and discovery features of ChatGPT, marking a key reversal for the firm. - The partnership with OpenAI could improve “licensing optics” and shift the narrative on the stock, according to analyst Mark Zgutowicz. - Getty shares were up 118% to $1.32 as of 12:44 p.m. in New York, putting them on track for the best session since July 2022. The stock had fallen about 55% this year to close at 61 cents on Thursday before the Juneteenth holiday weekend began. KOREA - SK Hynix - New #1 in South Korea: SK Hynix surpassed Samsung Electronics on Monday to become the country's most valuable listed company. - Remarkable turnaround: A striking reversal for a chipmaker that nearly collapsed under heavy debt roughly two decades ago. (CYCLES) - AI memory leader: Now the dominant supplier of high-bandwidth memory (HBM) chips powering AI systems. - Marquee customers: Key buyers include Nvidia (NVDA) and Alphabet's Google (GOOGL). - Massive 2026 rally: Shares are up more than 340% year-to-date, fueled by the global AI boom. - Market cap milestone: Valuation now exceeds both Samsung and Micron (MU). Markets Get Chopped - Questions being asked about if AI spend boom producing fast enough return - Back to earth on valuation scare - (all of a sudden?) - KOSPI down 11% - Chips getting hit - 12% for Memory ETF - MU down 9%, Intel 4%, ASML 7% RAM Prices... - Looking at some additional RAM today for some office computers .... --- ARE THEY KIDDING? RAM Prices Imminent Collapse???? - President Donald Trump said the prospect of global economic collapse was a big reason he signed an interim peace deal with Iran. - According to sources, the deal reopened the Strait of Hormuz and set in motion waivers for sanctions on Iran's oil sales to the international market, with the effect being an immediate drop in oil prices and a rise in US stocks. - The agreement has been seen as skewed in Iran's favor, giving the country broad gains before the next round of talks, and has prompted pushback and anger from Republican lawmakers. - MOU signed lat Wednesday - also now more waivers of sanctions on sale of Iranian oil - 60 day reprieve. China - Weak economic conditions - H Shares about to enter bear market - Hong Kong - Close to a technical bear market, dragged down by weak domestic consumption, a struggling property sector, and an exodus of funds fleeing "old tech" for AI plays elsewhere in Asia. - A-shares are listed in mainland China (Shanghai/Shenzhen) and primarily target domestic investors. H-shares are listed in Hong Kong and are freely available to international investors More China - Retail sales declined for the first time since December 2022, dropping 0.6% from a year earlier. - China's urban fixed-asset investment contracted 4.1% as of end-May, dragged by real estate and manufacturing. - Manufacturing fixed-asset investment contracted for the first time since December 2020. - Industrial output was the lone bright spot, rebounding from April's near three-year low. - The national unemployment rate fell to 5.1% in May, compared with 5.2% in April. Marrrr Jonggg - Mahjong can be highly addictive due to its rewarding blend of strategy, luck, and social interaction. The rapid tile-drawing, need for pattern recognition, and "just one more round" mentality trigger dopamine releases. If compulsive play disrupts your finances or daily life, it can become a behavioral addiction requiring intervention. - Tactile and Auditory Appeal: Many users on community forums like Reddit agree that the physical weight, texture, and distinct clinking sound of shuffling tiles provide soothing, sensory satisfaction. - There has been a 70% surge in mahjong content on TikTok in the past year - Yelp recently named the Chinese tile game a top trend of 2026, noting that searches for mahjong clubs surged 4,467% year over year for the period from September 2024 to August 2025 and that searches for mahjong lessons rose 819%. Alphabet - WHAT>????*&*^ - Alphabet shares slid 7%, on track for the search giant's worst day in a year. - Alphabet's Google has seen consecutive high-profile researchers leave in the last several days. - The company also has exposure to the market's concerns around commoditized AI and ballooning capital expenditures. - The share slide also came on the heels of a Sunday Wall Street Journal interview with Microsoft CEO Satya Nadella, who called for less dependence on “AI Giants” and said the AI market was commoditized. Back to Oracle - Oracle reduced workforce by 21,000 employees over past twelve months. - Cuts broader than previously disclosed, driven by artificial intelligence adoption. - Global headcount fell from 162,000 to 141,000 full-time employees year-over-year. - Workforce reductions generated $1.8 billion in restructuring costs, company reported. - Company warned AI deployment may continue resulting in workforce reductions. NVDA - Underperforming - Nvidia shares slipping recently despite remaining up about 12% in 2026. - Stock down roughly 3% past month, underperforming semiconductor peers. - SMH ETF surged 84% year-to-date, gaining 15% last month. - Traders predict Nvidia chip pricing power is beginning to decline. - Wall Street focus shifting toward memory and infrastructure AI buildout. - Micron and Sandisk shares jumped nearly 60% over past month. Gloom and Doom - JCD sent interesting take from Chris Bloomstran - Traditionally asset light companies with all sorts of revenue, high margins now.... ---- Converting into asset heavy with no real understanding of what the profitability or even revue will be in the future ----- Here are the highlights of his commentary we can explre: ------------AI buildout shifting markets from asset-light toward capital-intensive infrastructure cycle - Hyperscaler capex surge reflects move into heavy, long-duration asset base - Massive capital requirements challenge economics versus prior asset-light models - Depreciation burden rising sharply as infrastructure scales across AI ecosystem - Returns depend on utilization of expensive, long-lived physical compute assets - Asset-heavy cycles historically lead to overbuild, weak returns, eventual consolidation - Infrastructure spending absorbing nearly all operating cash flow for hyperscalers - Off-balance-sheet financing masking true scale of capital intensity shift - AI economics hinge more on physical capacity than software-driven scalability - Echoes of past asset-heavy booms with eventual oversupply and value destruction Amazon Day - Today - June 26th - US consumers will spend $26.3 billion online at Amazon and other retailers during the four-day sale, up 9% from last year's event in July, according to Adobe Inc. - About 201 million Amazon shoppers in the US were Prime subscribers as of March, up about 3% from a year earlier - Amazon will capture about 60% of all US online spending during Prime Day, its highest market share since 2019, according to estimates from EMarketer Inc. Chevron and Microsoft - Chevron Corp signed 20-year deal with Microsoft for data center power. - Agreement supplies natural-gas fired generation for massive West Texas facility. - Project Kilby expected online 2028, ramping to 2.67 gigawatts. - Full output enough to power more than 530,000 Texas homes. - Chevron partnering Engine No. 1, final investment decision planned later. - Deal follows prior reports of exclusive long-term power negotiations. More Oil News - Drill baby Drill - Interior Department cutting federal drilling bonds by 95% to spur exploration. - Required bond drops from $500,000 to $25,000 for leases. - Bonds ensure cleanup costs don't fall on taxpayers if wells abandoned. - Policy change aims to encourage more oil and gas development. - Proposal subject to 60-day public comment after Federal Register publication. FedEx Earnings - FedEx posted strong fiscal fourth-quarter earnings on Tuesday in the company's last quarter that included the freight business before its spin off. - FedEx Freight spun off into a separate publicly traded company on June 1. - The company said it saw a 3% year-over-year increase in domestic volume. - Stock down 6% A/H   Love the Show? Then how about a Donation? PayPal.Donation.Button({ env:'production', hosted_button_id:'JJJHP2GDEJC7J', image: { src:'https://www.paypalobjects.com/en_US/i/btn/btn_donateCC_LG.gif', alt:'Donate with PayPal button', title:'PayPal - The safer, easier way to pay online!', } }).render('#donate-button'); ANNOUNCING the THE CLOSEST TO THE PIN for SpaceX (SPCX) Winners will be getting great stuff like the new "OFFICIAL" DHUnplugged Shirt!     FED AND CRYPTO LIMERICKS   See this week's stock picks HERE Follow John C. Dvorak on Twitter Follow Andrew Horowitz on Twitter

Uno de los efectos colaterales de la fiebre por la inteligencia artificial es la carestía de la memoria RAM, un componente imprescindible en cualquier dispositivo electrónico de consumo, desde los ordenadores personales hasta las consolas de videojuegos pasando, naturalmente, por los teléfonos móviles. En los últimos dos años las grandes empresas tecnológicas se han lanzado a construir inmensos centros de datos para poder mover y entrenar gigantescos modelos de lenguaje. Todos esos servidores necesitan grandes cantidades de memoria, especialmente de un tipo muy avanzado conocido como HBM o memoria de alto ancho de banda, además de los módulos DDR5 más rápidos del mercado. El problema radica en que fabricar chips no es algo que se pueda acelerar de la noche a la mañana, la capacidad de la industria es limitada. Los tres principales fabricantes a nivel mundial, que son Samsung, SK Hynix y Micron, han visto que vender memoria para los servidores de inteligencia artificial es un negocio extremadamente rentable, mucho más que destinarlo a la electrónica de consumo. Por ello, han decidido desviar gran parte de sus líneas de producción hacia ese segmento tan lucrativo, lo que irremediablemente significa que están fabricando mucha menos memoria RAM tradicional para el mercado de consumo. Al haber mucha menos oferta de la memoria estándar en las tiendas y mantenerse la demanda, los precios se han disparado, han llegado a duplicarse o triplicarse desde finales del año pasado. Además del boom de los centros de datos, venimos arrastrando una situación creada por los propios fabricantes. Hace un par de años los precios de la memoria cayeron a mínimos históricos y estas empresas empezaron a perder dinero. Su reacción fue recortar la producción de forma intencionada para secar el mercado, eliminar el exceso de stock y recuperar sus márgenes de beneficio. Cuando quisieron darse cuenta, ese recorte premeditado se chocó de frente con la sed insaciable de chips de los gigantes de la inteligencia artificial. Todo junto ha creado gran escasez y la escalada de precios actual. No parece que los precios vayan a normalizarse a corto plazo. Montar una nueva fábrica de semiconductores para producir más chips cuesta miles de millones de euros y requiere años de planificación y construcción. Aunque la industria ya está invirtiendo en nuevas instalaciones, la mayor parte de esa capacidad de producción adicional no estará lista y operativa hasta el año 2027 o 2028, por lo que nos toca vivir una temporada con los precios bastante inflados. En medio de la tormenta está Apple, acostumbrada a exprimir su cadena de suministro, pero que ahora tendrá que subir los precios. Su problema es estructural, ya que contabiliza la memoria en el coste de los productos vendidos mientras los gigantes de la nube reparten ese gasto como inversión amortizable. La presión recae sobre unos márgenes que Wall Street espera que sigan subiendo. El daño va más allá, alcanza a todo el mercado del PC. Los analistas advierten que la escasez podría prolongarse como mínimo un par de años más. Mientras tanto, el usuario que renueva su móvil o su portátil estará financiando sin saberlo los servidores de la IA. En La ContraRéplica: 0:00 Introducción 3:40 Armagedón de la RAM 31:21 El pasaporte de Begoña 36:59 Las hijas de Zapatero 40:19 El voto CERA · Canal de Telegram: https://t.me/lacontracronica · “Contra el pesimismo”… https://amzn.to/4m1RX2R · “Hispanos. Breve historia de los pueblos de habla hispana”… https://amzn.to/428js1G · “La ContraHistoria del comunismo”… https://amzn.to/39QP2KE · “La ContraHistoria de España. Auge, caída y vuelta a empezar de un país en 28 episodios”… https://amzn.to/3kXcZ6i · “Contra la Revolución Francesa”… https://amzn.to/4aF0LpZ · “Lutero, Calvino y Trento, la Reforma que no fue”… https://amzn.to/3shKOlK Apoya La Contra en: · Patreon... https://www.patreon.com/diazvillanueva · iVoox... https://www.ivoox.com/podcast-contracronica_sq_f1267769_1.html · Paypal... https://www.paypal.me/diazvillanueva Sígueme en: · Web... https://diazvillanueva.com · Twitter... https://twitter.com/diazvillanueva · Facebook... https://www.facebook.com/fernandodiazvillanueva1/ · Instagram... https://www.instagram.com/diazvillanueva · Linkedin… https://www.linkedin.com/in/fernando-d%C3%ADaz-villanueva-7303865/ · Flickr... https://www.flickr.com/photos/147276463@N05/?/ · Pinterest... https://www.pinterest.com/fernandodiazvillanueva Encuentra mis libros en: · Amazon... https://www.amazon.es/Fernando-Diaz-Villanueva/e/B00J2ASBXM #FernandoDiazVillanueva #ram #ia Escucha el episodio completo en la app de iVoox, o descubre todo el catálogo de iVoox Originals

Semiconductors have moved from the background of the technology stack to the center of the AI economy. What used to be a specialized industry discussed mostly by engineers and investors is now shaping the speed, cost, and strategic direction of modern computing.In this episode of TechSurge, host Michael Marks speaks with Stacy Rasgon, Managing Director and Senior Analyst covering U.S. semiconductors and semiconductor capital equipment at Bernstein Research. Stacy has spent years analyzing the chip industry across cycles, but argues that the current moment feels different in scale: AI demand has created an unprecedented scramble for compute, memory pricing has surged, and companies across the stack are being forced to rethink capacity, architecture, and capital allocation.The conversation explains the 4 different kinds of semiconductor cycles—supply, inventory, product, and demand — and why Stacy believes the industry is currently in a demand cycle of unusual magnitude. The discussion also unpacks the distinction between DRAM and NAND, why high-bandwidth memory is becoming strategically central to AI systems, and how the physical realities of wafer capacity and silicon area are constraining supply in ways the broader market often misses.Stacy and Michael also discuss the hardware economics behind the current boom, with Michael pressing Stacy on why compute remains so scarce and how companies are improving performance through packaging and system design. Michael then moves the conversation beyond market headlines to the core business questions: who is actually paying for this compute, which use cases are generating real revenue, and whether AI spending is creating durable economic value or simply shifting costs elsewhere. Together, these questions highlight two of the episode's clearest insights: coding may be one of the earliest AI applications with meaningful willingness to pay, and inference, not training, is the real test of whether the current buildout becomes a lasting business or just another expensive wave of infrastructure.Stacy explains the concentration of power among the major wafer fabrication equipment players, the rise of ASICs as a meaningful share of AI silicon, Broadcom's rapidly expanding AI opportunity, and the growing role of Chinese companies as new entrants, especially in memory and semiconductor equipment. Along the way, the conversation asks the defining question facing the sector: is this just another semiconductor upswing, or the first true supercycle the industry has seen? Stacy believes that this might be the biggest supercycle he has seen in his career.Sign up for our newsletter at techsurgepodcast.com for updates on upcoming TechSurge Live Summits and future episodes.Links:Stacy Rasgon on LinkedIn: https://www.linkedin.com/in/stacy-rasgon-6924963Bernstein: https://www.alliancebernstein.com/corporate/en/home.htmlReferences Mentioned During the DiscussionNVIDIA Blackwell Platform: https://www.nvidia.com/en-us/data-center/blackwell-platform/High Bandwidth Memory (HBM) overview from Micron: https://www.micron.com/products/memory/hbmDRAM overview from IBM: https://www.ibm.com/think/topics/dramNAND flash overview from IBM: https://www.ibm.com/think/topics/nand-flash-memoryFurther ReadingMcKinsey on the semiconductor industry outlook: https://www.mckinsey.com/industries/semiconductors/our-insights/the-semiconductor-industry-in-2025Semiconductor Industry Association: 2025 State of the U.S. Semiconductor Industry: https://www.semiconductors.orgNVIDIA on the Blackwell architecture and AI infrastructure roadmap: https://www.nvidia.com/en-us/data-center/blackwell-platform/Broadcom AI investor materials and infrastructure commentary: https://investors.broadcom.comASML on lithography and advanced chip manufacturing: https://www.asml.com/en/technologyMicron on HBM and AI memory demand: https://www.micron.com/products/memory/hbmChapters[00:00:00] — Highlights[00:00:26] — Welcome to  the Episode[00:01:29] — Meet Stacy Rasgon[00:02:01] — Is This the First Real Semiconductor Supercycle?[00:05:33] — Inside the Strongest Memory Cycle in History [00:09:14] — Can Innovation Keep Up With AI Demand?[00:11:33] — Chiplets, Blackwell, and the New Economics of Compute [00:12:37] — What Could Signal the Cycle Is Slowing[00:14:26] — Vertical Integration at the Hyperscales [00:16:36] — The Difference between Apple and Meta[00:17:15] — What is Vertical Integration Being Done For?[00:18:15] — Will other bottlenecks develop as This Progresses? [00:21:13] — Oligopoly Pricing in the Market[00:22:22] — Any New Entrants into Memory?[00:23:46] — Why the Industry Must Pivot From Training to Inference[00:25:10] — Agentic Coding and the First Real AI Revenues[00:26:57] — Groq, Low-Latency Inference, and What GPUs Cannot Do Alone[00:29:28] —-Could The Smaller Companies All be Bought Up ?[00:30:19] — Why Semiconductor Equipment Matters More Than Ever [00:31:00] — How Semiconductor Equipment is Affected by the Cycle[00:32:55] — A Long Upcycle for Semiconductor Equipment Guys?[00:33:13] — The Big Five and the Rise of Chinese Equipment Players[00:34:24] — The Effects of Geopolitics[00:35:02] — Broadcom's Quiet AI Breakout[00:40:46] — ASICs vs GPUs and the Next Wave of Custom Chips[00:41:06] — Intel, Foundry Strategy, and the Long Turnaround[00:46:46] —-The Risks the Market May Still Be Underestimating[00:49:32] — Where Startups Still Have Room to Win[00:50:39] — What the Semiconductor Industry Could Look Like Next Year

李在明真的很沒見笑！ 韓國超荒謬，地方大選剛選完，選票不夠、投到半夜，整個「大選出包」，韓國選民紛紛要求學習台灣。結果超瞎總統李在明為了救股市，公開踩台灣一腳，真的讓人滿頭問號。 你一定覺得很奇怪，為什麼親中的李在明會贏得地方大選？其實不是韓國人親中，而是共同民主黨一路躺贏，背後真相看完本集就知道！ 黃仁勳空降首爾，不只替韓國科技股打氣，還直接上演「黃爸爸整頓韓國財閥」的大場面。平常高高在上的財閥大老，陪吃烤肉、陪發零食、陪蹲地合照，韓國社群直接傻眼！ 這集告訴你很多韓國的政經故事，竟然還跟台灣有點關係，歡迎分享給親友一起欣賞！ 全台獨家的世界經濟追劇深入報導，精彩萬分，持續連載中！ （現在就加入會員支持我們，還可以看到更多專屬影片～） https://www.youtube.com/@emmytw/join

La flambée actuelle des prix de la mémoire vive ne tombe pas du ciel. Elle est directement liée à l'explosion de l'intelligence artificielle. Les accélérateurs dédiés à l'IA, notamment les GPU utilisés dans les centres de données, consomment des quantités considérables de mémoire très rapide. Résultat : la demande dépasse l'offre, les prix montent, et une partie de l'industrie technologique se retrouve prise de court.Mais dans ce paysage sous tension, un acteur affirme avoir vu venir la crise : NVIDIA. Selon Collette Kress, directrice financière du groupe, l'entreprise avait anticipé la pénurie. Dans un entretien relayé par Wccftech, elle explique que NVIDIA « savait que cela allait arriver », contrairement à d'autres entreprises surprises par l'ampleur du phénomène. Pour elle, cette tension était prévisible, à condition de regarder suffisamment loin dans la chaîne d'approvisionnement.Pour comprendre l'enjeu, il faut revenir à la mémoire HBM, pour High Bandwidth Memory. Il s'agit d'une mémoire à très haute bande passante, conçue pour transférer énormément de données très rapidement entre les puces et les modèles d'IA. Elle est indispensable pour entraîner et faire fonctionner les grands modèles modernes. Chaque accélérateur peut embarquer des dizaines, voire des centaines de gigaoctets de cette mémoire ultra-rapide.Le problème, c'est que la production de HBM mobilise des ressources industrielles proches de celles utilisées pour fabriquer d'autres mémoires, comme la DDR présente dans les ordinateurs grand public. Quand l'IA absorbe une part croissante de ces capacités, le reste du marché se tend mécaniquement. Smartphones, PC, consoles ou composants grand public peuvent alors subir des hausses de prix. NVIDIA affirme avoir limité ce risque en passant commande très tôt. Mais le groupe ne s'est pas contenté d'acheter ce qui existait déjà. Collette Kress explique que l'entreprise travaille directement avec les trois grands fournisseurs de mémoire, en leur présentant ses futurs besoins et ses prochaines architectures. Autrement dit, NVIDIA ne subit pas seulement la chaîne d'approvisionnement : elle tente de la façonner en amont. Une stratégie qui illustre sa puissance actuelle. Dans la course à l'IA, le vainqueur n'est pas seulement celui qui conçoit les meilleures puces, mais aussi celui qui sécurise la mémoire nécessaire pour les faire tourner. Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.

The Head of our Europe and Asia Technology Team, Shawn Kim, explains how AI's appetite for memory chips is boosting the cost of everything from data centers to smartphones, with consequences that may reach far beyond the tech industry.Read more insights from Morgan Stanley.----- Transcript -----Shawn Kim: Welcome to Thoughts on the Market. I'm Shawn Kim, Head of Morgan Stanley's Europe and Asia Technology Team. Today, we're talking about chipflation – when memory chips stop getting cheaper over time, and become more expensive and even harder to find. It's Monday, June 8th, at 3pm in London.Memory chips are easy to ignore, until your laptop slows down, your phone costs more, or your cloud bill jumps. Memory is the computer's workspace. It holds whatever the machine needs at that moment, whether that is a web search, a video, a spreadsheet, or an AI model answering a question. DRAM is the fast memory inside servers, PCs and phones. NAND is what stores files in solid-state drives. And HBM, or high bandwidth memory, is the high-performance version sitting right next to the AI chip, helping them move huge amounts of data quickly. That last one – HBM – is key because AI has become intensely memory hungry. Memory prices have risen more than six-fold over the last year, a sharp break from decades when the cost of DRAM generally kept falling. The pressure is coming from AI infrastructure buildouts. We see servers accounting for 59 percent of DRAM demand by 2028, up from 37 percent in 2023. We also see enterprise solid-state drives reaching 65 percent of NAND demand, up from 18 percent. And simply put, data centers are taking a much bigger share of the memory pie. AI memory use is climbing fast, and at every scale. A newer AI chip uses 7.2 times more HBM than earlier generations. A full system uses about 65 times more. Across an entire AI data center buildout, the jump gets even bigger. HBM has gone from roughly 10 terabytes in 2020 to about 18 petabytes in 2026, orders of magnitude more. This demand is running into a supply chain that cannot respond quickly. New memory capacity takes years to build, qualify and ramp up. Supply relief is a process, not a switch. And that creates a two-tier market. Large AI and cloud buyers can sign long-term agreements, prepay and secure priority access. Traditional buyers, including PC makers, smartphone makers and industrial hardware companies, must compete for what remains. This impacts everyday products. In 2027, we see PC memory demand potentially facing a 15 percent shortfall, equivalent to about 58 million PCs. Smartphones could face a 12 percent shortfall, equivalent to about 134 million units. Companies may have to raise prices, cut specifications, delay launches, and accept lower profits. The dollar numbers are striking. We see the memory market growing from about $220 USD billion in 2025 to about $890 billion in 2026. Expectations for 2026 memory revenue rose 71 percent in just three months. That implies roughly $600 USD billion of incremental memory revenue in 2026, more than the annual market for smartphones, PCs, or servers, each taken on its own. The broader economy may not see a significant direct inflation shock. We estimate the direct impact on headline CPI at about 0.1 percent in 2026. But pressure is showing up in producer prices, in corporate margins, cloud costs, capital spending plans and delayed technology upgrades. AI has turned memory from the cheapest part of the digital economy into one of its most contested resources. These tiny chips most people never think of may now decide what gets built or delayed, and how much we all end up paying. Thanks for listening. If you enjoy the show, please leave us a review wherever you listen and share Thoughts on the Market with a friend or colleague today.

Nvidia unveiled the RTX Spark, an Arm-based consumer chip family built with MediaTek on TSMC 3, plus a DGX Station desktop that runs 1T-parameter models. Intel detailed its Crescent Island GPUs, MiniMax launched a coding model rivaling Opus 4.7 at 1/40th the price, and Anthropic bans AI in interviews. Nvidia announces the RTX Spark, an Arm-based consumer chip family it calls "the most efficient PC chip ever built", made on TSMC 3 in partnership with MediaTek (The Verge) Intel details its Crescent Island data center GPUs, built on its Xe3P architecture and using LPDDR5X memory instead of HBM, calling them "built for agentic AI" (Tom's Hardware) Nvidia unveils DGX Station for Windows, a desktop PC powered by a GB300 Grace Blackwell chip with up to 748 GB of memory, capable of running 1T-parameter models (SiliconAngle) Chinese AI developer MiniMax debuts M3, a new coding model that it says rivals Claude Opus 4.7, costing $0.12 per 1M input tokens, compared with $5 for Opus 4.7 (The Information) A look at Anthropic's hiring process, which prohibits AI use in interviews and features a culture interview that candidates describe as highly intense (Bloomberg) Learn more about your ad choices. Visit megaphone.fm/adchoices

Patrick Moorhead and Daniel Newman cover Daniel's acquisition of Enterprise Technology Research, IBM's historic $15 billion single-day commitment spanning quantum and open-source security, Anthropic's Claude Opus 4.8, and the heaviest single earnings night of the season featuring Dell, Marvell, Salesforce, Synopsys, Snowflake, HP, and Micron crossing $1 trillion in market cap. The handpicked topics for this week are: Anthropic Releases Claude Opus 4.8: Six Weeks After 4.7 Anthropic dropped Opus 4.8 just six weeks after 4.7, claiming it surpasses GPT-5.5 and Gemini 3.1 Pro on agentic coding, knowledge work, and computer use. Benchmark improvements across the board: agentic coding up from 64.3% to 69.2%, knowledge work from 1753 to 1890, agentic computer use from 82.8% to 83.4%. Three new features ship alongside it: Dynamic Workflows for multi-subagent orchestration inside Claude Code, Effort Control for managing token spend, and mid-task system messages via the API. Fast mode is now 2.5x faster and 3x cheaper. Pat's honest take: what it says on paper is good, particularly on tool triggering and citation precision, but he has lost significant trust in the company and is watching closely. (The Decode)   IBM Commits $10 Billion to Quantum: The Largest Single Quantum Bet in History IBM announced a $10 billion commitment over five years targeting a large-scale fault-tolerant quantum computer by 2029, landing the same day as the $5 billion Project Lightwell announcement for a single-day IBM strategic commitment of $15 billion. Pat has been calling 2029 to 2031 as the realistic commercial quantum window and calls this the strongest single corporate financial signal yet that the timeline is real. Daniel's framing: IBM wants to be the NVIDIA of quantum, and with a $10 billion commitment, it's sending a flare to the entire industry that pure-play quantum companies cannot compete at this balance sheet level. (The Decode)   IBM and Red Hat Launch Project Lightwell: $5B to Secure Open-Source Software IBM and Red Hat committed $5 billion and a global force of 20,000 engineers to secure open-source software for enterprises through frontier agentic AI, anchored by 11 of the largest US and Canadian banks including Bank of America, Goldman Sachs, JPMorgan Chase, Mastercard, and Visa. Pat's read: this is the productization answer to Anthropic Mythos. Mythos found the vulnerabilities. Lightwell is the industrial-scale patching and validation layer enterprises can actually buy on a subscription. Daniel adds that IBM is flexing its engineering talent base as a premium strategic asset, a direct counter to the narrative that AI replaces engineers. (The Decode)   Anthropic Project Glasswing: 23,000 Vulnerabilities Found Across 1,000 OSS Projects Anthropic's Claude Mythos scanned more than 1,000 widely deployed open-source projects and surfaced approximately 23,000 candidate vulnerabilities, with 1,094 confirmed as critical severity. The Cyber Verification Program now gates the strongest cyber-capable Claude variant behind vetted defenders only. While the tool creates real value, the surface of attack will likely grow as fast as any tool built to defend it. (The Decode)   Anthropic in Talks to Run Claude on Microsoft Maia 200 CNBC and The Information reported Microsoft is in active negotiations to supply Anthropic with its custom Maia 200 inference chip, which would make Anthropic the only frontier lab simultaneously running production workloads on four distinct silicon stacks: NVIDIA, AWS Trainium, Google TPU, and Microsoft Maia. Pat's context: Maia 200 delivers 30% better tokens per dollar than the latest Azure fleet per Satya Nadella, and this deal would be Maia's first major external deployment. Daniel's read: what can be built will be sold right now, and Anthropic chasing every available compute source is simply the structural reality of growing at 80x when you planned for 10x. (The Decode)   The Flip: Is AI CapEx Too Expensive to Earn Its Return? Pat takes the affirmative. With $725 billion in hyperscaler CapEx tracking for 2026, likely $1 trillion next year, memory has become the choke point making it even more expensive, and open-source models have closed enough of the quality gap for most enterprise tasks that the premium of frontier APIs is increasingly hard to justify. A recent Signal65 white paper shows on-prem payback at 18 months. Daniel's counter: Dell just booked $24 billion in AI orders in a single quarter. Agentforce crossed $1 billion ARR at 169% growth. NVIDIA guided to $91 billion. Only 20% of enterprises are using AI and only 2% of consumers. Both hosts admitted off the flip their notes looked nearly identical. (The Flip)   Micron Crosses $1 Trillion Market Cap Micron became the 12th US company ever to cross $1 trillion in market cap, surging 19% on May 26th as UBS raised its price target to $1,625, implying a $1.8 trillion market cap. Samsung's Q1 memory ASP jumped 146% year over year. DRAM spot prices spiked 55 to 60% quarter over quarter. Daniel has been pounding this call since sub-$100 and calls it a cycle elongated beyond anything seen in the 27 prior memory cycles, driven by HBM capacity reallocation away from consumer DRAM creating structural shortage. (Bulls and Bears)   Dell Technologies Q1 FY27: The Biggest Enterprise AI Infrastructure Print of 2026 Record $43.8 billion revenue, up 88% year over year, crushing the $35.7 billion consensus by $8 billion. AI-optimized servers at $16.1 billion, up 757% year over year. $24.4 billion in AI orders booked in a single quarter. FY27 AI server revenue guide raised from $50 billion to $60 billion. Non-GAAP EPS of $4.86 beat the $2.96 consensus by 64%. Stock up 18% after hours. Pat's framing: Dell was very clear about what they were going to do. Rack engineering, sales, and service. The basics. And they executed the basics at an extraordinary level while building a special relationship with NVIDIA who views Dell as a market maker for both enterprise and NeoCloud. Daniel's add: play nice and win. Michael Dell navigated the political landscape brilliantly and pulled the entire Dell brand along with him. (Bulls and Bears)   Marvell Technology Q1 FY27: Record Revenue, Data Center at 76% of Mix Record $2.418 billion revenue, up 28% year over year. Data center at $1.833 billion, up 27% year over year, now 76% of total revenue. Q2 guide of $2.7 billion at midpoint accelerates growth to 35% year over year. Operating cash flow a record $638.8 million. Daniel went on TV and said it's "written in the stars," arguing the market had misunderstood this one for too long by conflating its custom AI ASIC story with the full breadth of its connectivity and networking portfolio. Pat's closing: the shorts are eating it now and the custom AI ASIC versus merchant GPU debate is finally settling into the right answer, which is both in lockstep. (Bulls and Bears)   Salesforce Q1 FY27: Agentforce Crosses $1 Billion ARR Revenue $11.13 billion, up 13% year over year. Non-GAAP EPS of $3.88 crushed the $3.12 consensus by 24%. Agentforce ARR crossed $1 billion, up 169% year over year, with 28.6 trillion tokens processed, up 152% quarter over quarter. 50% of Agentforce bookings came from existing customers expanding. Daniel flagged the $25 billion accelerated buyback funded by new debt as an interesting signal worth watching. Pat's bottom line: it's not perfect, but certainly no "SaaSpocalypse" in those numbers. (Bulls and Bears)   Synopsys Q2 FY26: First Full Quarter With Ansys Integrated Revenue $2.276 billion, up 42% year over year, beating consensus. Non-GAAP EPS of $3.35 beat $3.15. FY26 guide raised to $9.665 billion midpoint. Daniel's framing: every chip runs through Synopsys tools, and the Ansys addition makes it the full-stack co-design platform Jensen Huang keeps talking about. Synopsys is not just the pick and shovel of current AI silicon. It is the pick and shovel of quantum, robotics, and space as well. (Bulls and Bears)   Snowflake Q1 FY27: Strongest Sequential Dollar Growth in Company History Product revenue $1.33 billion, up 34% year over year, the strongest sequential dollar growth in Snowflake history. Net revenue retention 126%. FY27 product revenue guide raised to $5.84 billion. Natoma acquisition announced for secure agentic enterprise connectivity. New $6 billion multi-year AWS commitment. Daniel's closing: proprietary unique data is the real moat of the agentic era, and that data has to live somewhere. It is going to go to platforms like Snowflake. (Bulls and Bears)   HP Inc. Q2 FY26: Eight Straight Quarters of Growth With AI PCs at 44% of Shipments Revenue $14.4 billion, up 9% year over year, the company marks its eighth consecutive quarter of top-line growth. Non-GAAP EPS of $0.86 beat the prior guide. Personal Systems at $10.2 billion, up 13%, with 30% operating profit growth. AI PCs jumped from 35% to 44% of shipments quarter over quarter, with HP guiding to 60 to 70% next fiscal year. FY26 EPS guide raised. Pat's note: they still need a permanent CEO, which would help investors sleep better at night. Daniel's add: the real explosive moment for device companies comes when AI moves to the edge and enterprises shift from expensive frontier model consumption to on-device inference. (Bulls and Bears)   Everpure Q1 FY27: Record Revenue, Rebrand Complete Record revenue of $1.1 billion, up 35% year over year. Product revenue $577 million, up 55%. Subscription ARR at $2 billion. FY27 guide raised to $4.41 to $4.51 billion. Pure Storage officially completed its rebrand to Everpure. Daniel's emerging thesis: the agentic era has focused enormous attention on memory and compute, but after the inference runs, the data has to sit somewhere. Storage has not seen its full inflection yet and Everpure is well positioned when that wave arrives. (Bulls and Bears)   The Decode Anthropic Releases Claude Opus 4.8 May 28  https://techcrunch.com/2026/05/28/anthropic-releases-opus-4-8-with-new-dynamic-workflow-tool/ IBM Commits $10B Over Five Years to Quantum Computing the Same Day as $5B Project Lightwell, Bringing IBM's One-Day AI https://www.barrons.com/articles/ibm-stock-quantum-computing-aafbb1eb IBM + Red Hat Announce Project Lightwell  https://newsroom.ibm.com/2026-05-28-ibm-and-red-hat-commit-5-billion-to-redefine-the-future-of-open-source-in-the-ai-era Anthropic Project Glasswing / Claude Mythos Finds 23,000 Potential Vulnerabilities Across 1,000+ Open-Source Projects https://www.securityweek.com/anthropic-mythos-detected-23000-potential-vulnerabilities-across-1000-oss-projects/ Anthropic Negotiating to Run Claude on Microsoft's Maia 200 AI Chips  https://www.cnbc.com/2026/05/21/anthropic-microsoft-maia-200-ai-chip.html OpenAI + Anthropic Walk Back the AI Jobs Apocalypse Ahead of IPOs https://finance.yahoo.com/sectors/technology/articles/ai-chiefs-walk-back-job-193605798.html https://x.com/RiskCentre/status/2059397756016611668 The Flip Is AI Capex Becoming Too Expensive to Earn Its Return — and Will the Result Be a Forced Shift to Open-Source and Smaller Use-Case-Specific Models, or a Continued $725B+ Hyperscaler Buildout That Vindicates the Capex on Productivity Gains? FOR:  The shift is to open-source + smaller use-case-specific models with better token economics, not away from AI https://x.com/danielnewmanUV/status/2059822712122400975 DeepSeek 75% permanent price cut + Anthropic Claude Code restriction reversal https://www.buildfastwithai.com/blogs/ai-news-today-may-26-2026 $190B Microsoft capex + $725B+ aggregate hyperscaler capex with no analog ROI yet  https://www.buildfastwithai.com/blogs/ai-news-today-may-26-2026   AGAINST:  Salesforce Agentforce ARR crossed $1B this quarter on 28.6T tokens processed  https://www.stocktitan.net/sec-filings/CRM/8-k-salesforce-inc-reports-material-event-3b8ead2852bb.html Lenovo +105% AI revenue, +84% Q4; Dell $43B AI backlog: the AI infrastructure flywheel is converting capex to revenue today https://investor.marvell.com/news-events/press-releases/detail/1023/marvell-technology-inc-reports-first-quarter-of-fiscal-year-2027-financial-results NVIDIA $91B Q2 guide + $1T Blackwell+Vera Rubin CY25-CY27 reaffirmed  https://www.cnbc.com/2026/05/20/were-raising-our-price-target-on-nvidia-after-another-knockout-quarter-and-guide-.html DeepSeek + Chinese price war is a Chinese export-controls story, not a US economic ceiling story https://www.cnbc.com/2026/05/21/anthropic-microsoft-maia-200-ai-chip.html   Bulls & Bears Micron (NASDAQ: MU) Crosses $1 TRILLION Market Cap for the First Time https://www.cnbc.com/2026/05/26/micron-stock-trillion-market-cap.html Dell Technologies Q1 FY27 ACTUALS  https://www.cnbc.com/2026/05/28/dell-q1-earnings-report-2027.html Marvell Technology Q1 FY27 ACTUALS https://investor.marvell.com/news-events/press-releases/detail/1023/marvell-technology-inc-reports-first-quarter-of-fiscal-year-2027-financial-results Salesforce CRM Q1 FY27 ACTUALS  https://investor.salesforce.com/financials/quarterly-results/ Synopsys SNPS Q2 FY26 ACTUALS https://investor.synopsys.com/events-and-presentations/events/event-details/2026/Q2-Fiscal-Year-2026-Earnings/default.aspx Snowflake SNOW Q1 FY27 ACTUALS  https://www.businesswire.com/news/home/20260527027931/en/Snowflake-Reports-Financial-Results-for-the-First-Quarter-of-Fiscal-2027 HP Inc. HPQ Q2 FY26 ACTUALS https://finance.yahoo.com/markets/stocks/articles/hp-q2-earnings-call-highlights-230459161.html Everpure (NYSE: P, formerly Pure Storage) Q1 FY27 ACTUALS  https://investor.salesforce.com/financials/quarterly-results/ Synopsys SNPS Q2 FY26 ACTUALS https://investor.synopsys.com/events-and-presentations/events/event-details/2026/Q2-Fiscal-Year-2026-Earnings/default.aspx Snowflake SNOW Q1 FY27 ACTUALS  https://www.businesswire.com/news/home/20260527027931/en/Snowflake-Reports-Financial-Results-for-the-First-Quarter-of-Fiscal-2027 HP Inc. HPQ Q2 FY26 ACTUALS  https://finance.yahoo.com/markets/stocks/articles/hp-q2-earnings-call-highlights-230459161.html Everpure (NYSE: P, formerly Pure Storage) Q1 FY27 ACTUALS https://www.prnewswire.com/news-releases/everpure-announces-first-quarter-fiscal-2027-financial-results-302783502.html

Nvidia entra en la guerra del portátil con chips Arm propios, Intel promete hardware para IA agéntica sin HBM, Vast se convierte en unicornio 3D, un gran sindicato docente pide frenar la IA en primaria y los astrónomos detectan una posible fábrica de planetas más allá de Júpiter.Puedes seguirnos en YouTube en https://youtube.com/olivernabani y puedes unirte al Discord Mashain en https://olivernabani.com/discord

Dans ce nouvel épisode de Trends Bourse, Gaële Poncelet et Guy Legrand décryptent les signaux contradictoires qui agitent les marchés financiers. Le PIB américain du premier trimestre a été révisé à la baisse, tandis que la confiance des consommateurs américains continue de se dégrader. Faut-il y voir un signal d'alarme pour l'économie mondiale ou simplement une inquiétude passagère liée à l'inflation et au prix des carburants ? Pendant ce temps, deux nouveaux géants technologiques viennent de franchir le seuil symbolique des 1.000 milliards de dollars de valorisation : Micron et SK Hynix. Longtemps considérés comme de simples fournisseurs de composants, ces spécialistes des puces mémoire HBM sont devenus des acteurs incontournables de la révolution de l'intelligence artificielle. Un épisode indispensable pour comprendre les nouvelles dynamiques des marchés, les enjeux de l'intelligence artificielle et les opportunités qui se cachent au-delà des géants traditionnels de la tech. Trends Bourse est une chaîne podcast de Trends-Tendances. Plus d'informations et de conseils pour vos investissements sur www.tendances.be/bourse. Vous désirez recevoir chaque jour des conseils d'investissements dans votre boîte électronique, enregistrez-vous gratuitement sur www.tendances.be/newsletters.   Hosted by Simplecast, an AdsWizz company. See https://pcm.adswizz.com for information about our collection and use of personal data for advertising.

Sat, 30 May 2026 08:17:00 +0200https://mario-lochner-weil-dein-geld-mehr-kann.blogs.audiorella.com/412-new-episode64673de7-5434-4422-88a6-5818be332bf2Teste Shopify kostenlos und bring' dein Geschäft auf das nächste Level – auf https://www.shopify.com/de/marioGemacht für Deutschland, powered by Shopify --> https://www.shopify.com/de/marioKostenlos auf wikifolio registrieren:https://www.wikifolio.com/de/de/home?utm_source=mariolochner&utm_content=homeZum wikifolio “Rohstoffwerte” von Thomas Dellmann:https://www.wikifolio.com/de/de/w/wf00trohst?utm_source=mariolochner&utm_content=dellmannDer Aktienmarkt jagt von Rekord zu Rekord – doch warum eigentlich?In dieser Ausgabe von „Das Briefing“ analysieren wir die entscheidenden Themen für deutsche Privatanleger: Iran-Krieg, Trump, fallender Ölpreis, neue Rekorde an den Börsen, die nächste Phase der KI-Rally, Drohnen-Aktien, Inflation, Zinsen und die möglichen Mega-IPOs von SpaceX, Anthropic und OpenAI.Die Märkte setzen offenbar auf Entspannung im Iran-Konflikt. Trump bleibt kryptisch, doch die Spekulationen über ein Memorandum of Understanding nehmen zu – und der Ölpreis fällt deutlich. Ist das der nächste Treiber für Aktien? Oder unterschätzt der Markt die Risiken?Gleichzeitig läuft die KI-Rally weiter. Micron, SK Hynix und der gesamte HBM- und Halbleiter-Sektor stehen im Fokus. Welche Aktien profitieren noch vom KI-Boom? Wo ist die Rally schon heiß gelaufen – und wo gibt es noch Chancen?Außerdem schauen wir auf einen neuen Favoritenbereich: Drohnen-Aktien. Die US-Regierung könnte in diesen Sektor einsteigen – und genau solche politischen Signale können an der Börse massive Bewegungen auslösen. Bei Dell konnte man gerade sehen, wie stark solche Entwicklungen wirken können. Hatte Trump hier wieder den richtigen Riecher?Und dann geht es um Leopold Aschenbrenner, eines der spannendsten KI-Genies der Welt. Er setzt unter anderem auf Nebius und CleanSpark – aber das sind nicht seine einzigen Favoriten. Welche Aktien könnten von der nächsten KI-Welle profitieren?Dazu gibt es einen wichtigen AHA-Effekt bei Inflation und Zinsen. Viele Anleger schauen gerade auf die falschen Signale. Warum könnte der Markt hier auf dem falschen Fuß erwischt werden? Und was bedeutet das für Aktien, Anleihen, Tech-Werte, Nasdaq, S&P 500 und den Dax?Zum Schluss blicken wir auf mögliche Mega-Börsengänge von SpaceX, Anthropic und OpenAI. Sind solche IPOs ein Warnsignal wie in früheren Marktphasen? Oder ist das erst der Beginn der nächsten großen Rally?In dieser Folge geht es um:✅ Aktienmarkt auf Rekordhoch: Warum steigen DAX, S&P 500 und Nasdaq weiter?✅ Iran-Krieg, Trump und Ölpreis: Kommt jetzt Entspannung?✅ KI-Aktien: Micron, SK Hynix, Nebius, CleanSpark und weitere Favoriten✅ Drohnen-Aktien: Neuer Boom durch mögliche Unterstützung der US-Regierung?✅ Inflation und Zinsen: Das übersehene Risiko für Anleger✅ SpaceX, Anthropic und OpenAI: Warnsignal oder Start der nächsten Börsen-Euphorie?✅ Welche Aktien jetzt noch Potenzial haben könnten„Das Briefing“ ist das wöchentliche Börsenformat für deutsche Privatanleger – jeden Samstag mit Analyse, Einordnung und den wichtigsten Themen für dein Depot.TIME STAMPS(00:00) Intro(00:40) Aufreger: Zauberformel Wegnehmen(02:44) BRIEFING – Watchlist(04:53) Bären: Inflation und Gier glühen(07:39) Bullen: Kurz vor Deal, Ölpreis fällt(11:45) Geldidee spezial (Werbung)(14:35) DOSSIER: SpaceX – das bedeuten Mega-IPOs(18:52) MINDBLOW: Das wird überraschen(21:00) Steigt S&P 500 auf 9.000 Punkte?(23:28) GELDIDEEN: Nachzügler, Drohnen, AschenbrennerWeil dein Geld mehr kann! Und du auch!AKTIEN, ETFs & GELDIDEEN: Auf diesem Kanal wollen wir vor allem darüber reden, wie Du mehr aus deinem Geld machen kannst. Deswegen zeige ich dir die besten Geldideen unter Aktien & ETFs. Natürlich reden wir auch über Bitcoin und ImmobilienINTERVIEWS: Dazu spreche ich mit den besten Experten der Finanzszene, mit Fondsmanagern, Vermögensverwaltern und mit erfolgreichen Unternehmern – beispielsweise sind regelmäßig zu Gast auf meinem Kanal: Dr. Andreas Beck, Dr. Gerd Kommer, Martin Hackler, Christian Fuchs, Mojmir Hlinka, Hendrik Leber und Philipp Vorndran–––––––––––––––––––––––––––––––––––––––––––––Wichtig: Dieser Podcast dient nur zur Information und ist KEINE Anlageberatung! Es handelt sich nicht um Empfehlungen und die Betreiber dieses Kanals haften nicht für etwaige Verluste, die sich aus Investments ergeben, die aufgrund von Informationen aus diesem Podcast getätigt werden. Investieren birgt Risiken, handle verantwortungsbewusst.* Hierbei handelt es sich um einen Affiliate-Link. Wenn du auf diesen Link klickst und etwas abschließt, erhalten wir (je nach Anbieter) eine Provision. Dir entstehen dadurch keine Mehrkosten und du unterstützt unser Projekt. Wir danken dir für deinen Support!fullfalseerfolg,aktien,wirtschaft,börse,geld,motivation,investieren,etfs,investments

Dan Nathan hosts Dan Niles of Niles Investment Management on the Risk Reversal podcast to discuss macro conditions, AI-driven market leadership, and lessons from prior tech cycles. Niles compares the current AI build-out to 1997–1998's internet infrastructure boom, arguing recent macro scares (tariffs, Iran/oil) created buying opportunities and that a bubble can persist, with further gains likely before a potential 30–50% drawdown next year. He cites a January 30 “agentic AI” step-change increasing token/compute demand, supporting strong CapEx and earnings growth, and notes Nvidia's growth versus valuation relative to past leaders like Cisco. They debate rising yields, inflation measures, and expectations for a rate-cutting Fed chair (Kevin Warsh). The conversation covers Intel's potential benefit from agentic shifts, corporate AI cost pressures, likely disruption to software/IT services and knowledge work, Micron's HBM-driven surge and cyclicality risks, and how major IPOs like SpaceX, OpenAI, and Anthropic could reshape flows and create new short opportunities. —FOLLOW USYouTube: @RiskReversalMediaInstagram: @riskreversalmediaTwitter: @RiskReversalLinkedIn: RiskReversal Media

就在本周，全球三大存储芯片巨头——SK 海力士、三星、美光——市值同时突破万亿美元。过去一年，SK 海力士股价累计涨幅超过 900%。财报显示，SK 海力士 2026 年第一季度营收接近三倍于去年同期，三星存储部门同期利润同比暴增 8 倍。存储，正在成为这一轮 AI 浪潮中最炙手可热的硬件赛道。 这背后是一场史无前例的供需错配。SK 海力士 2026 年全年 HBM 产能已经售罄，短缺预计将持续至 2027 年。与此同时，AI 服务器对内存的疯狂抢占，已经传导到了普通消费者：同等配置的手机更贵了，想要的机型根本没货——内存，被 AI 抢走了。但存储行业是出了名的「涨得多猛、跌得多惨」——过去 40 年，一个周期内暴涨 10 倍、暴跌 90% 的情况屡见不鲜。这一轮，究竟是又一场超级周期，还是 AI 真的重写了游戏规则？ 这期节目我们请来前半导体行业从业者、公众号「傅立叶的猫」主理人张海军，我们与他聊了聊：为什么存储行业现在如此火爆、这一轮周期在哪些维度上真的不一样；从 HBM 缺货到 NAND 角色转变，存储的五层架构如何被 AI 重塑；以及，国内长鑫、长江存储的进展，和周期拐点究竟何时到来。 本期人物 Yaxian，「科技早知道」主播 张海军，前新思科技高级工程师、半导体公众号「傅里叶的猫」主理人 时间轴 [02:46] 这轮周期是什么时候开始的？ 去年 Q1 现货价格已现端倪，三星海力士宣布停产 DDR4 引发囤货恐慌 三重触发因素叠加：AI 需求拉动、DDR4 停产、关税预期 [04:05] 存储的五层架构：离处理器越近，速度越快、价格越贵 从 HBM、DRAM 到 NAND、机械硬盘，带宽与价格的权衡逻辑 "内存墙"：芯片算力已不是瓶颈，存储带宽才是 [07:33] HBM 为什么缺货缺到 2027 年？ 英伟达每代芯片 HBM 需求翻倍 台积电 CoWoS 先进封装扩产周期长，供给端是硬约束 [09:26] 三家原厂格局：海力士领跑，三星和美光追赶 HBM 市场从 350 亿美元预计增至 2027 年近 800 亿美元 第四代 HBM 格局重洗 [13:00] 这轮为什么不一样：结构性转变的五个维度 需求端：从周期性补库存转向 AI 结构性爆发，三重需求（DRAM、HBM、NAND）同时拉升 供给端：原厂从抢份额转向利润优先，扩产极为克制； 长协模式（LTA）：锁量不锁价，甚至要求客户绑定原厂资本开支 [22:37] 看多 vs 看空：分歧在哪里？ 短期超买、股价涨幅过大是看空方主要依据 需求缺口太大、算法效率提升反而会扩大需求 [25:48] 绕过 HBM 瓶颈的各路方案：英伟达 CMX、谷歌 CXL 内存池、Cerebras 晶圆级芯片 英伟达 CMX 方案用 DPU 预取 NAND 数据降低延迟 谷歌用 CXL 协议构建 TB 级 DRAM 内存池 Cerebras 晶圆芯片面临散热、扩展性硬伤 [32:00] NAND 的角色转变：从"温数据仓库"到 AI 推理参与者 AI agent 推理链路长、中间结果需要落盘，NAND 直接参与推理过程 闪迪 HBF（高带宽闪存）：容量是 HBM 的 8~16 倍、成本更低，但受限于擦写寿命和工作温度 [38:06] 长鑫、长江存储即将上市，国内存储进展如何？ 长江存储 NAND 国内已大规模销售，长鑫 HBM 尚未量产商用 小米等手机厂商参股长鑫，本质是提前锁定产能 阶跃星辰 Step 3.7 Flash 模型 阶跃星辰 Step 3.7 Flash 是专门面向生产级 Agent 的高效率 Flash 模型，一开始就为 Agent、Coding、Search 和多模态工作流设计，让你的生产工作流又快又稳。 感兴趣的小伙伴欢迎点击链接（ https://sourl.co/ZUaCXf ）试试，新注册用户有代金券，可以直接上手跑一跑，看看它在你的工作流里表现怎么样。 幕后制作 监制：Yaxian 后期：迪卡 运营：George 设计：饭团 商业合作 声动活泼商业化小队，点击链接直达声动商务会客厅（https://sourl.cn/9h28kj ），也可发送邮件至 business@shengfm.cn 联系我们。 加入声动活泼 声动活泼正在招聘全职商务运营经理、早咖啡内容实习生和社群实习生，如果你也对播客行业的内容制作感兴趣，欢迎点击招聘入口 关于声动活泼 「用声音碰撞世界」，声动活泼致力于为人们提供源源不断的思考养料。 我们还有这些播客：声动早咖啡、声东击西、吃喝玩乐了不起、反潮流俱乐部、泡腾 VC、商业WHY酱、跳进兔子洞 、不止金钱 欢迎在即刻、微博等社交媒体上与我们互动，搜索 声动活泼 即可找到我们。 期待你给我们写邮件，邮箱地址是：ting@sheng.fm 欢迎扫码添加声小音，在节目之外和我们保持联系。Special Guest: 张海军.

Nesse vídeo analisamos o desempenho de gigantes como Nvidia, Intel e Micron, e como o boom de semicondutores e chips HBM impacta o mercado financeiro global. Entenda a correlação entre o investimento em IA e o crescimento econômico de Taiwan e Coreia do Sul, comparando o cenário atual com a crise das empresas "ponto com" nos anos 2000. Avaliamos riscos, projeções de lucratividade e o impacto macroeconômico real desse avanço tecnológico. Descubra se estamos vivendo uma nova bolha da inteligência artificial.00:00:00 – Analisando a atual bolha da IA00:00:59 – Desempenho dos índices de mercado financeiro00:01:58 – Nvidia e empresas de hardware beneficiadas00:03:24 – O papel estratégico dos chips HBM00:04:33 – Concentração excessiva no índice S&P 50000:05:42 – Boom da IA na Ásia impactante00:09:24 – Impactos na economia global e PIB00:10:30 – Exportações de Taiwan e Coreia disparando00:15:10 – Comparação com bolhas históricas do mercado00:22:00 – Conclusão sobre riscos e investimentos futuros

00:00 - Issues with Singh's wearing patke03:15 - Technique for tying a patka 04:45 - Joora and receding hairlines08:40 - Canadian content creators 10:50 - Why did HBM change their content and move away from brown topics?14:40 - Indy's not a fan of Shai Gilgeous-Alexander (SGA)15:50 - Growing with your audience 17:20 - What type of content does best?20:30 - Knowing comes from doing22:10 - People need to understand there's enough to go around 24:30 - Being at the Nagar Kirtan in Surrey28:20 - Apologising for lack of awareness30:30 - Condemning Parmvsthewrld's actions31:50 - Rules were given to N3ON; he just didn't follow it35:33 - Sikh policing in Canada is more extreme than in the UK 40:10 - Everyone's journey with Sikhi is different 43:40 - What's going on with the Indian hate movement in Canada?45:30 - Do immigrants need to integrate?48:50 - Going into drive-thrus on a tractor52:10 - Celebrity houses being robbed in Canada 53:30 - Crossbow Singh - the best meme55:26 - Why is the Khalistan movement so strong in Canada? 57:45 - What would actually happen if Khalistan existed?01:02:55 - Quick-fire questions01:05:03 - HBM  dropping final thoughtsFollow Hours Before Midnight:https://www.youtube.com/@UCwQd_2VrNwysI0il5dmdnXg Follow Ekdeep: https://www.instagram.com/ekdeep21k/Follow Us On:TikTok - https://bit.ly/indy-and-dr-tik-tokInstagram - http://bit.ly/indy-and-dr-instaFacebook - http://bit.ly/indy-and-dr-facebookSpotify - http://bit.ly/indy-and-drAlso available at all podcasting outlets.#hoursbeforemidnight #canadianpolitics #sikh #indianculture #desiculture

En el Radar Empresarial de hoy ponemos la atención en SK Hynix, que, igual que Micron, ha logrado entrar en el grupo de compañías valoradas en más de un billón de dólares. La empresa surcoreana se ha beneficiado del impulso de la inteligencia artificial y de las mejores previsiones de firmas financieras. Ese contexto explica el avance en el Kospi, donde protagonizó una de las subidas de 2026. Este año, sus acciones acumulan un crecimiento cercano al 250% aproximadamente. La compañía se ha convertido en una pieza esencial dentro del sector de memorias, considerado de los próximos meses. Una de las razones principales es su estrecha relación con Nvidia, ya que es único proveedor de memorias HBM para la tecnológica estadounidense. Estos componentes permiten gestionar volúmenes de datos con velocidad, algo imprescindible para los desarrollos vinculados a la inteligencia artificial. Con ese liderazgo, SK Hynix controla el 57% del mercado de HBM y el 32% del segmento DRAM. El aumento de la demanda de semiconductores y la escasez de suministros han elevado la importancia estratégica de fabricantes como SK Hynix En apenas dieciséis meses, la empresa pasó de 100.000 millones de dólares a superar el billón de capitalización Sin embargo, la compañía reconoce los riesgos en la industria Su presidente, Chey Tae-won, advirtió hace meses que la falta de obleas de silicio podría provocar un desequilibrio hasta 2030 y generar una diferencia entre oferta y demanda 20%. Las advertencias de SK Hynix coinciden con las de otros directivos del sector tecnológico Michael Dell aseguró que la demanda de memoria inteligencia artificial podría multiplicarse por 625 durante años el director ejecutivo de Micron, Sanjay Mehrotra, explicó en CNBC que la memoria se ha convertido en un recurso imprescindible Según el ejecutivo, los sistemas de inteligencia artificial necesitan más capacidad y un rendimiento superior para desarrollar todo su potencial y sostener el crecimiento esperado del sector tecnológico

La course à l'intelligence artificielle entre dans une nouvelle dimension. OpenAI annonce vouloir porter sa capacité de calcul à 30 gigawatts d'ici 2030. Pour donner un ordre de grandeur, un gigawatt correspond à la puissance d'un réacteur nucléaire. Autrement dit, l'objectif équivaut à plusieurs dizaines de centrales électriques mobilisées pour faire tourner des systèmes d'IA.Aujourd'hui, l'entreprise dispose d'environ 1,9 gigawatt. Elle vise donc une multiplication par seize en cinq ans. Une montée en puissance spectaculaire, portée par le succès de ses services depuis le lancement de ChatGPT et par une demande mondiale en forte croissance. Mais OpenAI n'est pas seule. Amazon et Anthropic ont eux aussi annoncé des investissements massifs, avec plusieurs gigawatts de capacité en préparation. La compétition est désormais industrielle.Pour atteindre ces objectifs, il faudra des infrastructures colossales : centres de données, réseaux électriques renforcés, et surtout des composants électroniques très spécialisés. OpenAI travaille notamment sur une puce maison intégrant de la mémoire HBM — une technologie ultra-rapide empilée en couches, essentielle pour traiter d'énormes volumes de données. Problème : cette mémoire est aujourd'hui rare. Les fabricants comme Samsung ou SK Hynix peinent à suivre la demande. Cette tension pourrait avoir des répercussions concrètes : hausse des prix pour les ordinateurs, les smartphones ou les consoles, faute de composants disponibles.Mais l'enjeu dépasse l'économie. Il est aussi environnemental. Alimenter 30 gigawatts de calcul implique une consommation énergétique massive, sans parler du refroidissement des serveurs, qui nécessite souvent d'importantes quantités d'eau. Si cette énergie n'est pas décarbonée, l'empreinte carbone de l'IA pourrait fortement augmenter. Le secteur fait donc face à un dilemme : soutenir une innovation technologique majeure, tout en limitant son impact écologique. Certaines entreprises explorent déjà des solutions, comme l'utilisation d'énergies renouvelables ou l'optimisation des algorithmes pour consommer moins. Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.

Le patron de Mistral.ai dresse un tableau sans concession de l'IA en Europe • Google muscle Android avec Gemini et réinvente la souris • TikTok attaqué en France • Un conflit social chez Samsung menace la production mondiale de puces • Les rédactions bousculées par l'IA • Publier un livre avec l'intelligence artificielle • Un futur moteur de recherche français boosté à l'IA • Gare à l'IA documentaire mal maîtrisée en entreprise. ⭐️ Découvrez Frogans, l'innovation française qui réinvente le Web [PARTENARIAT]===============================Plaidoyer sans concession pour la souveraineté européenne de l'IAAuditionné à l'Assemblée nationale, Arthur Mensch, cofondateur et directeur général de Mistral AI, a livré une analyse offensive sur l'état de l'intelligence artificielle en Europe. Pour lui, l'IA est devenue une ressource stratégique comparable à l'énergie : elle conditionne la souveraineté économique, militaire et culturelle du continent. Il appelle à investir massivement dans les modèles, les infrastructures et l'électricité bas carbone, faute de quoi l'Europe risque de dépendre durablement des États-Unis et de la Chine. Reste à savoir si Mistral aura les moyens financiers et politiques de rivaliser avec les hyperscalers américains.Vidéo de l'audition d'Arthur Mensch : https://www.youtube.com/watch?v=kKWOkWv6pJMLa cybersécurité nouveau champ de bataille IALa course à l'IA se déplace sur le terrain stratégique de la cybersécurité. OpenAI aurait présenté un modèle spécialisé capable d'anticiper des vulnérabilités inédites, avec un accès encadré mais ouvert à certaines organisations européennes, tandis que Mistral AI travaillerait également sur un modèle dédié. Ces outils, capables de détecter des failles avant qu'elles ne soient exploitées, deviennent des instruments de souveraineté numérique. Leur contrôle et leur accès sont désormais des enjeux diplomatiques autant que technologiques.Android passe à l'IA agentiqueLors de sa conférence Android, Google a présenté une version agentique de Gemini appelée à s'intégrer au cœur d'Android. Capable d'agir à la place de l'utilisateur, l'assistant pourra naviguer dans les applications, extraire des informations ou remplir des formulaires, avec validation humaine pour les actions sensibles. Une évolution majeure vers un smartphone proactif, qui transforme l'IA en véritable copilote numérique. Le déploiement est attendu progressivement sur les futurs appareils Android.Google réinvente la souris avec l'IALa filiale DeepMind de Google a dévoilé un prototype baptisé “Magic Pointer”, combinant capture d'écran locale et intelligence artificielle. En survolant un contenu, l'utilisateur peut demander instantanément un graphique, un résumé ou un recalcul contextuel. Cette interaction homme-machine repensée pourrait intégrer Chrome et les Chromebooks à terme. Une démonstration spectaculaire qui illustre l'intégration toujours plus fine de l'IA dans les gestes informatiques du quotidien.TikTok dans le viseur de familles françaises endeuilléesSeize familles réunies au sein du collectif Algos Victima ont déposé plainte contre TikTok pour abus de faiblesse, après plusieurs suicides d'adolescentes. Elles accusent l'algorithme de la plateforme d'avoir favorisé l'exposition répétée à des contenus anxiogènes ou dangereux. Au-delà du volet judiciaire, l'affaire relance le débat sur la régulation des réseaux sociaux et sur une possible interdiction avant 15 ans.La CNIL alerte sur les lunettes connectéesLa CNIL (Commission Nationale Informatique et Libertés) met en garde contre les risques de surveillance diffuse liés aux lunettes équipées de caméras et de micros. L'autorité française appelle à un usage responsable et propose plusieurs recommandations pour préserver la vie privée. Si ces dispositifs peuvent rendre des services, notamment pour les personnes malvoyantes, ils posent une question sociétale majeure : comment éviter une banalisation de la captation d'images dans l'espace public ?Samsung sous tension : menace sur la production mondiale de puces IAEn Corée du Sud, 50 000 salariés de Samsung menacent de se mettre en grève pour réclamer une meilleure redistribution des bénéfices liés à l'IA. Le groupe est l'un des rares fabricants mondiaux de mémoire HBM, composant clé des puces d'intelligence artificielle. Un arrêt prolongé de la production pourrait perturber toute la chaîne mondiale des semi-conducteurs. La planète tech observe avec inquiétude ce bras de fer social.IA et journalisme : vers la fin des tâches répétitives ?En direct de Gaspésie, Bruno Guglielminetti, animateur du podcast Mon Carnet, analyse l'impact de l'IA dans les rédactions. Entre menaces de grève et suppression de postes, la réécriture automatisée des dépêches cristallise les tensions. Pour lui, l'IA peut devenir un assistant éditorial précieux, libérant du temps pour l'enquête et le terrain. Un basculement culturel qui oblige les écoles de journalisme à revoir leurs formations.Mon nouveau livre sur le podcasting... publié grâce à l'IAJe vous présente mon livre Lancez votre podcast à l'ère de l'IA, consacré à la création de podcasts avec les outils d'intelligence artificielle. Autoédité avec l'appui d'outils IA pour la structuration, la correction et la mise en forme, le livre explore les nouvelles pratiques du média audio. Une démonstration concrète des mutations en cours dans l'édition et la création de contenus.Ibou, le pari d'un moteur de recherche françaisSylvain Peyronnet, cofondateur et CEO d'Ibou, présente ce moteur de recherche conversationnel français, conçu pour proposer une information plus transparente et pluraliste. L'objectif : éviter les bulles de filtrage et valoriser les sources en exposant les différents points de vue. Grâce aux modèles de langage récents, l'équipe affirme pouvoir bâtir un moteur performant sans dépendre d'une collecte massive de données utilisateurs. Un défi ambitieux face aux géants américains.IA et gestion documentaire : les risques invisibles en entreprise [PARTENARIAT]Guillaume Brault, directeur technique Europe du Sud chez Box, alerte sur les dangers d'une IA branchée sur des bases documentaires mal gouvernées. Sans classification préalable, un agent conversationnel peut faire remonter des informations sensibles à des collaborateurs non habilités. La clé réside dans l'étiquetage et la gouvernance des données afin de contrôler précisément ce que les modèles peuvent consulter et restituer. L'IA devient ainsi un révélateur des failles organisationnelles internes.Hébergé par Audiomeans. Visitez audiomeans.fr/politique-de-confidentialite pour plus d'informations.

真的假的？川普要輝達蘋果特斯拉轉單英特爾，對台積電和台灣反而會更好？ 我們是在自我安慰還是自爽？都不是，本集含金量高到哭，讓我們研究團隊完整破解給你看！ 台積電股價創新高，但現在川普力挺英特爾18A搶台積電客戶，輝達蘋果特斯拉通通要轉單了嗎？三星也虎視眈眈，護國神山領先地位岌岌可危？18A有這麼厲害嗎？ 半導體爭霸戰現在究竟誰佔上風？AMD竟是記憶體大缺貨的最大得利者，英特爾最大的敵人不是TSMC而是蘇姿丰蘇媽？台灣連線如何撿到世界最大的一把槍？ Emmy本集一次解答，絕對精采，千萬別錯過！請把這集分享給你每一個關心台積電、台灣半導體和科技產業的朋友！ 全台獨家的世界經濟追劇深入報導，精彩萬分，持續連載中！ （現在就加入會員支持我們，還可以看到更多專屬影片～） https://www.youtube.com/@emmytw/join

In this episode, Felicity Thomas and Candice Bourke sit down with Armina Rosenberg, Co-Founder of Minotaur Capital, for a fascinating deep dive into the next wave of AI investing opportunities, global equity themes and the market trends sophisticated investors are watching closely right now.The conversation explores why Minotaur Capital believes the biggest opportunities in AI may now sit beyond Nvidia particularly across AI infrastructure, memory semiconductors, power networks and the “picks and shovels” powering the AI revolution.Armina shares:Why SK Hynix is Minotaur's highest conviction global stock ideaWhy the memory super cycle could still be in its early stagesWhether Nvidia is still attractive at current valuationsWhy Minotaur remains short TeslaThe risks around overcrowded AI trades and hyperscaler spendingOpportunities in defence, energy transition and AI infrastructureHow Minotaur uses proprietary AI tools internally to analyse markets faster than traditional fund managersThe growing disconnect between market headlines and underlying fundamentalsWhat could trigger a meaningful AI-fuelled correction in marketsThe discussion also unpacks some incredible real-world examples of how AI is already changing investing, including Minotaur's internally developed AI research system Taurient, which helps the team analyse global companies, reporting seasons and thematic opportunities at extraordinary speed and scale.Armina also explains why memory semiconductors particularly HBM memory may become one of the most important and profitable areas of the global technology ecosystem over the coming years, driven by explosive demand from AI models, hyperscalers and data centres.If you're interested in AI investing, Nvidia, semiconductors, SK Hynix, Tesla, global equities, thematic investing or the future of financial markets this episode is packed with institutional grade insights and investment ideas.

In KI-Beschleunigern steckt der superschnelle Spezialspeicher HBM – und wir erklären, wie er funktioniert: Folge 2026/9 des Podcasts Bit-Rauschen.

2330不只是台積電的代碼，也是股價！ 未來的日子裡，還是回台灣爽笑韓國哭嗎？ 台股勢如破竹站上四萬點！DRAM持續缺貨中，三星記憶體營收獲利大爆發，會威脅到護國神山嗎？魏哲家該擔心嗎？ Sk海力士為了擊敗三星，竟想找TSMC代工HBM4E？韓媒踢爆三星二奈米製程良率膨風，馬斯克找三星代工AI晶片是請鬼拿藥單嗎？ AI時代的半導體爭霸戰，台積電三星英特爾馬斯克TeraFab誰將勝出？分享影片給你的親友吧！ 全台獨家的世界經濟追劇深入報導，精彩萬分，持續連載中！ （現在就加入會員支持我們，還可以看到更多專屬影片～） https://www.youtube.com/@emmytw/join 臺日茶交流

Many investors were not aware of Advantest. This Japanese company quietly controls roughly 70% of the global semiconductor test equipment market — the quality assurance layer that every AI chip, every HBM memory module, and every packaged GPU must pass through before it ships to a customer. As AI chips have gotten more complex and more expensive, the cost of shipping a faulty one has risen dramatically, and the demand for Advantest's equipment has followed.The stock reflects that. Up 450% in the past year. Up roughly 250% since CSI first wrote about Advantest eleven months ago. Analyst earnings per share expectations have roughly doubled in twelve months — that is what drove the run. Advantest just reported FY2025 results of $7.1 billion in revenue and guided FY2026 to approximately $9 billion, a 26% year-over-year increase. In a test equipment market the company itself sizes at $12.5 to $13 billion for 2026, that would put Advantest's global market share approaching 70% — a level of dominance that is genuinely rare in any industry.Nick and Kasey cover the full picture in CSI's first public video on Advantest: how it became the undisputed leader, why the test equipment slice of the $140 billion wafer fab equipment market is small but critical, what the 47x forward earnings and 56x forward free cash flow multiples actually imply, and why some analysts are already flagging a potential cycle downturn starting in 2027 even as the bulls hold firm.The close is pure CSI. Radical moderation. Patience is a strategy. Stay in the game and survive.What we cover:— Advantest FY2025: $7.1B revenue and dominant market share vs. Teradyne and Aehr— FY2026 guidance: ~$9B — approaching 70% of a $12.5–13B global TAM— Why AI chips and packaged modules require testing — and why it is getting more expensive— Test equipment as a slice of the $140B WFE pie — small, critical, and cyclical— Valuation: 47x forward P/E and 56x forward FCF — what the re-rate means now— Analyst EPS doubled in twelve months — the mechanics of why the stock ran— The 2027 cycle risk: bear vs. bull analyst expectations laid out clearly— Radical moderation: the CSI framework for parabolic stocks and surviving the cycleSponsored by fiscal.ai — the platform behind CSI's research charts. Get 15% off at fiscal.ai/csiDisclosure: This content is for general information only and is not individual investment advice. All investing involves risk.chipstockinvestor.com

Stocks for Beginners and Tykr proudly present "Weekend Watchlist". We dissect a company using Tykr's risk rating and fair value analysis process. Learn how to avoid emotional mistakes, choose investments with a rationale, and build wealth with confidence. Get your free trial and special discount offer. Join Tykr today and take advantage of this special offer of 30% off with coupon code SAVE30. See for yourself why Tykr is the essential tool for every serious DIY share investor. 14-day free trial included, then a no-quibble 30-day money back guarantee: Get your free trial and special discount offer.Micron Technology has become a critical player in the AI revolution - not because it makes GPUs, but because it makes the high‑bandwidth memory that keeps those GPUs fed with data. In this episode, we explore Micron's business model, why HBM is so essential, and how Micron has become responsible for more than half of all earnings‑estimate upgrades across the S&P 500.Weekend Watchlist is about helping beginners sharpen their investing process through real companies and real stories.Disclosure: The links provided are affiliate links. I will be paid a commission if you use this link to make a purchase. You will receive a discount by using these links/coupon codes. I only recommend products and services that I use and trust myself or where I have interviewed and/or met the founders and have assured myself that they're offering something of value. Stocks for Beginners is a production of Finpods Pty Ltd. The advice shared on Stocks for Beginners is general in nature and does not consider your individual circumstances. Opinions expressed by guests are theirs alone and may not represent the views of Finpods, Money Sherpa, or Phil Muscatello. Stocks for Beginners exists purely for educational and entertainment purposes and should not be relied upon to make an investment or financial decision. If you do choose to buy a financial product, read the PDS, TMD, and obtain appropriate financial advice tailored towards your needs. Philip Muscatello and Finpods Pty Ltd are authorised representatives of Money Sherpa PTY LTD ABN - 321649 27708, AFSL - 451289. Hosted on Acast. See acast.com/privacy for more information.

Shares for Beginners and Tykr proudly present "Weekend Watchlist". We dissect a company using Tykr's risk rating and fair value analysis process. Learn how to avoid emotional mistakes, choose investments with a rationale, and build wealth with confidence. Get your free trial and special discount offer. Join Tykr today and take advantage of this special offer of 30% off with coupon code SAVE30. See for yourself why Tykr is the essential tool for every serious DIY share investor. 14-day free trial included, then a no-quibble 30-day money back guarantee: Get your free trial and special discount offer.Micron Technology has become a critical player in the AI revolution - not because it makes GPUs, but because it makes the high‑bandwidth memory that keeps those GPUs fed with data. In this episode, we explore Micron's business model, why HBM is so essential, and how Micron has become responsible for more than half of all earnings‑estimate upgrades across the S&P 500.Weekend Watchlist is about helping beginners sharpen their investing process through real companies and real stories.Disclosure: The links provided are affiliate links. I will be paid a commission if you use this link to make a purchase. You will receive a discount by using these links/coupon codes. I only recommend products and services that I use and trust myself or where I have interviewed and/or met the founders and have assured myself that they're offering something of value. Shares for Beginners is a production of Finpods Pty Ltd. The advice shared on Shares for Beginners is general in nature and does not consider your individual circumstances. Opinions expressed by guests are theirs alone and may not represent the views of Finpods, Money Sherpa, or Phil Muscatello. Shares for Beginners exists purely for educational and entertainment purposes and should not be relied upon to make an investment or financial decision. If you do choose to buy a financial product, read the PDS, TMD, and obtain appropriate financial advice tailored towards your needs. Philip Muscatello and Finpods Pty Ltd are authorised representatives of Money Sherpa PTY LTD ABN - 321649 27708, AFSL - 451289. Hosted on Acast. See acast.com/privacy for more information.

芯片突围有多难？3D堆叠如何引爆HBM革命？国产键合设备凭什么比海外快一倍？从材料到封装，大咖蔡维伽揭秘半导体硬核实战。02:03 HBM封装线：行业挑战与趋势的深入剖析04:03 人生态度：在创业压力与意义上寻找平衡06:08 材料与工艺的完美匹配：国产设备在键合领域的挑战与机遇08:34 挑战与机会：国产化设备在微细加工中的应用前景11:24 设备厂商的挑战：如何实现关键设备的清洗和消毒？14:15 HBM技术：通讯速度更快、运算效率更高的突破方案17:08 三D堆叠技术：异构集成的泛型三D堆叠领域19:59 国产半导体厂商的挑战与机遇：在巨头林立的行业中的生存之道22:49 后发优势与供应链自主可控：国产设备制造业的现状与前景25:43 临时晶合胶：公平竞争的供应链选择与技术进步28:32 纳米颗粒的挑战与客户需求：理解技术与实现全球最好的解决方案31:26 AI时代下的职场挑战与机遇：保持兴趣与学习的重要性合作洽谈添加微信: xinpianjiemi01(添加请备注：粉丝)发布平台：微信公众号|喜马拉雅|小宇宙|微博|知乎|雪球|搜狐网|bilibili|今日头条|视频号|支付宝|抖音|快手|小红书欢迎粉丝们积极在评论区和我们留言互动哦，同时欢迎大家提出你们最想知道的芯片问题，优质提问将有机会得到产业大咖一对一解答！千万别错过~

Our 239th episode with a summary and discussion of last week's big AI news!FYI: this one has pretty out of date news, I was traveling last week and failed to upload... apologies. Recorded on 03/25/2026Hosted by Andrey Kurenkov and Jeremie HarrisFeel free to email us your questions and feedback at andreyvkurenkov@gmail.com and/or hello@gladstone.aiRead out our text newsletter and comment on the podcast at https://lastweekin.ai/In this episode:OpenAI is discontinuing the Sora iPhone app and seemingly shutting down its video generation API, while retaining internal video world-modeling work; the move is framed as a compute- and focus-driven pivot toward coding and productivity agents, alongside a collapsed Disney Sora deal. Anthropic's Claude Code/Cowork gains full computer control via keyboard/mouse/display, tied to the recent Cept acquisition, and Google's Gemini rolls out background “task automation” on select phones for limited delivery/ride-share use. Cursor releases the cheaper, benchmark-strong Composer 2 coding model amid controversy over its Kimi-based origins and licensing attribution. Other items include Adobe Firefly custom model training, Luma's Uni 1 image model, US contracting and legislative proposals affecting AI safeguards and state preemption, major chip/memory developments (Meta ASICs with Broadcom, Micron's HBM-driven surge, Musk's “Terra Fab”), robotaxi scaling, and research on monitoring agent misalignment, shutdown resistance, “consciousness cluster” preferences, and self-improving “hyper agents.”Timestamps:(00:00:10) Intro / BanterTools & Apps(00:01:48) OpenAI Discontinues Sora App, Shuts Down Video Generation Service and API - Bloomberg(00:07:12) Anthropic's Claude Code and Cowork can control your computer | The Verge(00:13:15) Gemini task automation is slow, clunky, and super impressive | The Verge(00:19:44) Cursor Launches Composer 2 AI Model to Challenge OpenAI & Anthropic(00:28:28) Adobe's AI image generator can now be trained on your own art | The Verge(00:29:40) Luma AI launches Uni-1, a model that outscores Google and OpenAI while costing up to 30 percent less | VentureBeatApplications & Business(00:32:41) Trump Contracting Clause Would Override AI Safeguards(00:40:00) Meta accelerates AI ASIC roll-out as Broadcom secures four-generation chip design deal(00:47:07) Micron revenue almost triples, tops estimates as demand for memory soars(00:50:54) Elon Musk Unwraps $25 Billion Terafab Chip-Building Project - CNET(00:56:40) Zoox to widen US robotaxi footprint with San Francisco, Vegas expansion(00:57:39) Waymo hits 170 million miles while avoiding serious mayhem | The VergePolicy & Safety(00:58:43) The White House just laid out how it wants to regulate AI | CNN Business(01:06:54) How we monitor internal coding agents for misalignment(01:12:30) Incomplete Tasks Induce Shutdown Resistance in Some Frontier LLMs(01:18:15) Summary: Mechanisms to Verify International Agreements about AI Development(01:23:09) Scoop: Anthropic meets with House Homeland Security behind closed doorsResearch & Advancements(01:24:24) Consciousness Cluster: Preferences of Models that Claim they are Conscious(01:30:22) HyperAgentsSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

[심층 인터뷰] HBM 다음은? 터보퀀트 등장과 반도체 생존 전략

Simon Erickson of 7investing breaks down two high-potential watchlist stocks: Micron Technology (NASDAQ:MU) — a memory giant riding the AI boom with explosive margin expansion — and Infleqtion (NYSE:INFQ), a newly public quantum computing company using groundbreaking neutral atom technology. Micron's high bandwidth memory (HBM) is completely sold out through 2026, with gross margins expanding 16 percentage points year-over-year on $13.6B in quarterly revenue. This is one of the most compelling AI infrastructure plays in the semiconductor space right now.Infleqtion just hit public markets via SPAC in February 2026 and is already partnered with NVIDIA (NASDAQ:NVDA) through a CUDA integration called Qlink. With quantum computing threatening RSA encryption and unlocking solutions classical computers can't touch, government contracts, defense spending, and research grants are flooding the space — justifying premium valuations for early-stage leaders.

삼테성즈 2026년 3월호. 아이들 감시용 지피티를 직접 만들다? 그리고 HBM 칩의 신묘한 냉각 시스템!-오프닝비둘기 뇌를 해킹해 만드는 러시아 생체 트론한국 유튜버, 일론 머스크의 뉴럴링크 실험 지원- 이용의 디벼보기아이들을 감시하기 위해 지피티를 직접 만들어 보다!?- K2의 공학썰HBM칩은 도대체 어떻게 냉각하는거냐-자료https://www.slideshare.net/slideshow/2626-3-k2-hbm-pdf/286682771과학과사람들 제공

男人的狀態，不是年紀的問題，是你有沒有訓練。野馬波15分鐘，非侵入、無恢復期，不靠藥、不硬撐，而是讓身體，自己找回力量。我是張英傑醫師輝達泌尿科院長想了解更多,歡迎諮詢 https://fstry.pse.is/8vpytd —— 以上為 Firstory Podcast 廣告 —— 日前 SK 海力士與晟碟聯合舉辦HBF啟動儀式，期望成為 AI 推論應用記憶體新主流，主要還是想要解決 HBM 供給吃緊的窘境，卻也引發市場的高度議論，真能取代 HBM嗎？《時間軸》00:00 開場00:30 市場動態04:47 什麼是 HBF ? 與 HBM 有何不同 ?12:13 各界對 HBF 的看法以及台廠的機會19:02 留言 QA【財訊Podcast】聽更多： https://open.firstory.me/user/wealth1...原文刊登於財訊雙週刊 759 期文章連結 https://www.wealth.com.tw/articles/13...主持：陳碧珠來賓：楊喻斐錄音：蔡克承成音：蔡克承錄音日期：2026.03.24

Is High-Bandwidth Flash (HBF) a direct competitor to HBM, or something entirely new? In this episode, we break down the technology of High-Bandwidth Flash (HBF) and its role in solving the "memory wall" for AI inference.We explore how HBF fits into the existing memory hierarchy—sitting between HBM and traditional NAND flash—to provide high capacity and increased speed for the next generation of AI data centers. We also discuss the technical side of the SanDisk and Kioxia partnership, including BiCS technology and CBA wafer bonding, and when investors can expect this technology to actually hit the bottom line.Join us on Discord with Semiconductor Insider, sign up on our website: www.chipstockinvestor.com/membershipSupercharge your analysis with AI! Get 15% of your membership with our special link here: https://fiscal.ai/csi/Sign Up For Our Newsletter: https://mailchi.mp/b1228c12f284/sign-up-landing-page-short-formChapters:01:17 – The NAND Market: Key Players (Micron, Samsung, SK hynix) 02:10 – What is NAND Flash? Memory vs. Storage 02:50 – The Memory Hierarchy: Capacity vs. Speed 03:57 – What is High-Bandwidth Flash (HBF)? 04:32 – HBF vs. HBM: Clearing up the Misunderstandings 05:10 – The Tech: BiCS, 16-Chip Stacks, and Wafer Bonding 05:57 – Solving the AI "Memory Wall" & Inference Bottleneck 06:58 – Timeline: When Will HBF Generate Revenue? 08:01 – Investment Strategy: The Fab Five & Lam ResearchIf you found this video useful, please make sure to like and subscribe!*********************************************************Affiliate links that are sprinkled in throughout this video. If something catches your eye and you decide to buy it, we might earn a little coffee money. Thanks for helping us (Kasey) fuel our caffeine addiction!Content in this video is for general information or entertainment only and is not specific or individual investment advice. Forecasts and information presented may not develop as predicted and there is no guarantee any strategies presented will be successful. All investing involves risk, and you could lose some or all of your principal. #HBF #HighBandwidthFlash #HBM #AIInvesting #Semiconductors #ChipStockInvestor #SanDisk #NAND #DataCenter #AIHardwareNick and Kasey own shares of Sandisk

Hoy en 10 Minutos con Sammy: Anthropic captura el 73% del gasto en IA de nuevas empresas tras negarse a quitar barreras de seguridad al Pentágono. Micron casi triplica ingresos gracias a la memoria HBM que devoran los chips de IA. Perplexity lanza gratis su navegador Comet en iOS — y reconoce que recopila tus datos para publicidad. Un satélite del tamaño de una caja de zapatos busca vida en 50.000 millones de planetas con IA a bordo. Y un agente de IA de Meta provoca un incidente Sev 1 dejando datos de usuarios expuestos durante dos horas — mientras a la directora de seguridad otro agente le borró todo el inbox.Puedes seguirnos en YouTube en https://youtube.com/olivernabani y puedes unirte al Discord Mashain en https://olivernabani.com/discord

Onto Innovation is a critical "choke point" in the semiconductor supply chain, providing the essential metrology and inspection gear that makes advanced packaging and high-bandwidth memory (HBM) possible. In this deep dive, we break down why their flagship Dragonfly system is a winner in the AI era and how the company is successfully pivoting growth toward Taiwan and South Korea to offset shifts in China.We also take a close look at Onto's rivalry with industry giant KLA Corp and how their recent Semi Lab acquisition is already fueling a massive ramp-up in 2026 revenue. With free cash flow per share exploding over the last five years, we're explaining why this mid-cap power player has earned its spot as a core long-term holding in our portfolio.Join us on Discord with Semiconductor Insider, sign up on our website: www.chipstockinvestor.com/membershipSupercharge your analysis with AI! Get 15% of your membership with our special link here: https://fiscal.ai/csi/Sign Up For Our Newsletter: https://mailchi.mp/b1228c12f284/sign-up-landing-page-short-formChapters:00:00 The "Fab Five" vs. Onto Innovation 02:22 Metrology, Yield, and Defect Control 03:26 Front-End vs. Advanced Packaging03:32 What is the Dragonfly? Optical Metrology Explained 04:20 High Bandwidth Memory (HBM) & GPU Integration 04:49 Financial Turnaround: Revenue & 2026 Forecasts 05:19 The Semi Lab Acquisition Impact 05:54 Geographic Shift: China vs. Taiwan & South Korea 07:55 Free Cash Flow Growth & Portfolio Status 08:38 Risks: Mind the Industry CycleIf you found this video useful, please make sure to like and subscribe!*********************************************************Affiliate links that are sprinkled in throughout this video. If something catches your eye and you decide to buy it, we might earn a little coffee money. Thanks for helping us (Kasey) fuel our caffeine addiction!Content in this video is for general information or entertainment only and is not specific or individual investment advice. Forecasts and information presented may not develop as predicted and there is no guarantee any strategies presented will be successful. All investing involves risk, and you could lose some or all of your principal.#Semiconductors #Investing #OntoInnovation #AdvancedPackaging #AI #StockMarket #ChipStockInvestor #Metrology #HBMNick and Kasey own shares of Onto Innovation

SUMMARY DEL SHOW Futuros en verde tras una semana débil por shock de energía, pero el tape sigue dominado por Irán y el Estrecho de Ormuz Crudo volátil con Brent cerca de $106 y WTI alrededor de $96, mientras Trump presiona a aliados para reabrir rutas de envío $NBIS se dispara por acuerdo de infraestructura de IA con $META, $MU acelera capacidad en Taiwán para DRAM y HBM, y $BABA prepara un AI agent empresarial sobre Qwen

Join Kyle, Nader, Vibhu, and swyx live at NVIDIA GTC next week!Now that AIE Europe tix are ~sold out, our attention turns to Miami and World's Fair!The definitive AI Accelerator chip company has more than 10xed this AI Summer:And is now a $4.4 trillion megacorp… that is somehow still moving like a startup. We are blessed to have a unique relationship with our first ever NVIDIA guests: Kyle Kranen who gave a great inference keynote at the first World's Fair and is one of the leading architects of NVIDIA Dynamo (a Datacenter scale inference framework supporting SGLang, TRT-LLM, vLLM), and Nader Khalil, a friend of swyx from our days in Celo in The Arena, who has been drawing developers at GTC since before they were even a glimmer in the eye of NVIDIA:Nader discusses how NVIDIA Brev has drastically reduced the barriers to entry for developers to get a top of the line GPU up and running, and Kyle explains NVIDIA Dynamo as a data center scale inference engine that optimizes serving by scaling out, leveraging techniques like prefill/decode disaggregation, scheduling, and Kubernetes-based orchestration, framed around cost, latency, and quality tradeoffs. We also dive into Jensen's “SOL” (Speed of Light) first-principles urgency concept, long-context limits and model/hardware co-design, internal model APIs (https://build.nvidia.com), and upcoming Dynamo and agent sessions at GTC.Full Video pod on YouTubeTimestamps00:00 Agent Security Basics00:39 Podcast Welcome and Guests07:19 Acquisition and DevEx Shift13:48 SOL Culture and Dynamo Setup27:38 Why Scale Out Wins29:02 Scale Up Limits Explained30:24 From Laptop to Multi Node33:07 Cost Quality Latency Tradeoffs38:42 Disaggregation Prefill vs Decode41:05 Kubernetes Scaling with Grove43:20 Context Length and Co Design57:34 Security Meets Agents58:01 Agent Permissions Model59:10 Build Nvidia Inference Gateway01:01:52 Hackathons And Autonomy Dreams01:10:26 Local GPUs And Scaling Inference01:15:31 Long Running Agents And SF ReflectionsTranscriptAgent Security BasicsNader: Agents can do three things. They can access your files, they can access the internet, and then now they can write custom code and execute it. You literally only let an agent do two of those three things. If you can access your files and you can write custom code, you don't want internet access because that's one to see full vulnerability, right?If you have access to internet and your file system, you should know the full scope of what that agent's capable of doing. Otherwise, now we can get injected or something that can happen. And so that's a lot of what we've been thinking about is like, you know, how do we both enable this because it's clearly the future.But then also, you know, what, what are these enforcement points that we can start to like protect?swyx: All right.Podcast Welcome and Guestsswyx: Welcome to the Lean Space podcast in the Chromo studio. Welcome to all the guests here. Uh, we are back with our guest host Viu. Welcome. Good to have you back. And our friends, uh, Netter and Kyle from Nvidia. Welcome.Kyle: Yeah, thanks for having us.swyx: Yeah, thank you. Actually, I don't even know your titles.Uh, I know you're like architect something of Dynamo.Kyle: Yeah. I, I'm one of the engineering leaders [00:01:00] and a architects of Dynamo.swyx: And you're director of something and developers, developer tech.Nader: Yeah.swyx: You're the developers, developers, developers guy at nvidia,Nader: open source agent marketing, brev,swyx: and likeNader: Devrel tools and stuff.swyx: Yeah. BeenNader: the focus.swyx: And we're, we're kind of recording this ahead of Nvidia, GTC, which is coming to town, uh, again, uh, or taking over town, uh, which, uh, which we'll all be at. Um, and we'll talk a little bit about your sessions and stuff. Yeah.Nader: We're super excited for it.GTC Booth Stunt Storiesswyx: One of my favorite memories for Nader, like you always do like marketing stunts and like while you were at Rev, you like had this surfboard that you like, went down to GTC with and like, NA Nvidia apparently, like did so much that they bought you.Like what, what was that like? What was that?Nader: Yeah. Yeah, we, we, um. Our logo was a chaka. We, we, uh, we were always just kind of like trying to keep true to who we were. I think, you know, some stuff, startups, you're like trying to pretend that you're a bigger, more mature company than you are. And it was actually Evan Conrad from SF Compute who was just like, you guys are like previousswyx: guest.Yeah.Nader: Amazing. Oh, really? Amazing. Yeah. He was just like, guys, you're two dudes in the room. Why are you [00:02:00] pretending that you're not? Uh, and so then we were like, okay, let's make the logo a shaka. We brought surfboards to our booth to GTC and the energy was great. Yeah. Some palm trees too. They,Kyle: they actually poked out over like the, the walls so you could, you could see the bread booth.Oh, that's so funny. AndNader: no one else,Kyle: just from very far away.Nader: Oh, so you remember it backKyle: then? Yeah I remember it pre-acquisition. I was like, oh, those guys look cool,Nader: dude. That makes sense. ‘cause uh, we, so we signed up really last minute, and so we had the last booth. It was all the way in the corner. And so I was, I was worried that no one was gonna come.So that's why we had like the palm trees. We really came in with the surfboards. We even had one of our investors bring her dog and then she was just like walking the dog around to try to like, bring energy towards our booth. Yeah.swyx: Steph.Kyle: Yeah. Yeah, she's the best,swyx: you know, as a conference organizer, I love that.Right? Like, it's like everyone who sponsors a conference comes, does their booth. They're like, we are changing the future of ai or something, some generic b******t and like, no, like actually try to stand out, make it fun, right? And people still remember it after three years.Nader: Yeah. Yeah. You know what's so funny?I'll, I'll send, I'll give you this clip if you wanna, if you wanna add it [00:03:00] in, but, uh, my wife was at the time fiance, she was in medical school and she came to help us. ‘cause it was like a big moment for us. And so we, we bought this cricket, it's like a vinyl, like a vinyl, uh, printer. ‘cause like, how else are we gonna label the surfboard?So, we got a surfboard, luckily was able to purchase that on the company card. We got a cricket and it was just like fine tuning for enterprises or something like that, that we put on the. On the surfboard and it's 1:00 AM the day before we go to GTC. She's helping me put these like vinyl stickers on.And she goes, you son of, she's like, if you pull this off, you son of a b***h. And so, uh, right. Pretty much after the acquisition, I stitched that with the mag music acquisition. I sent it to our family group chat. Ohswyx: Yeah. No, well, she, she made a good choice there. Was that like basically the origin story for Launchable is that we, it was, and maybe we should explain what Brev is andNader: Yeah.Yeah. Uh, I mean, brev is just, it's a developer tool that makes it really easy to get a GPU. So we connect a bunch of different GPU sources. So the basics of it is like, how quickly can we SSH you into a G, into a GPU and whenever we would talk to users, they wanted A GPU. They wanted an A 100. And if you go to like any cloud [00:04:00] provisioning page, usually it's like three pages of forms or in the forms somewhere there's a dropdown.And in the dropdown there's some weird code that you know to translate to an A 100. And I remember just thinking like. Every time someone says they want an A 100, like the piece of text that they're telling me that they want is like, stuffed away in the corner. Yeah. And so we were like, what if the biggest piece of text was what the user's asking for?And so when you go to Brev, it's just big GPU chips with the type that you want withswyx: beautiful animations that you worked on pre, like pre you can, like, now you can just prompt it. But back in the day. Yeah. Yeah. Those were handcraft, handcrafted artisanal code.Nader: Yeah. I was actually really proud of that because, uh, it was an, i I made it in Figma.Yeah. And then I found, I was like really struggling to figure out how to turn it from like Figma to react. So what it actually is, is just an SVG and I, I have all the styles and so when you change the chip, whether it's like active or not it changes the SVG code and that somehow like renders like, looks like it's animating, but it, we just had the transition slow, but it's just like the, a JavaScript function to change the like underlying SVG.Yeah. And that was how I ended up like figuring out how to move it from from Figma. But yeah, that's Art Artisan. [00:05:00]Kyle: Speaking of marketing stunts though, he actually used those SVGs. Or kind of use those SVGs to make these cards.Nader: Oh yeah. LikeKyle: a GPU gift card Yes. That he handed out everywhere. That was actually my first impression of thatNader: one.Yeah,swyx: yeah, yeah.Nader: Yeah.swyx: I think I still have one of them.Nader: They look great.Kyle: Yeah.Nader: I have a ton of them still actually in our garage, which just, they don't have labels. We should honestly like bring, bring them back. But, um, I found this old printing press here, actually just around the corner on Ven ness. And it's a third generation San Francisco shop.And so I come in an excited startup founder trying to like, and they just have this crazy old machinery and I'm in awe. ‘cause the the whole building is so physical. Like you're seeing these machines, they have like pedals to like move these saws and whatever. I don't know what this machinery is, but I saw all three generations.Like there's like the grandpa, the father and the son, and the son was like, around my age. Well,swyx: it's like a holy, holy trinity.Nader: It's funny because we, so I just took the same SVG and we just like printed it and it's foil printing, so they make a a, a mold. That's like an inverse of like the A 100 and then they put the foil on it [00:06:00] and then they press it into the paper.And I remember once we got them, he was like, Hey, don't forget about us. You know, I guess like early Apple and Cisco's first business cards were all made there. And so he was like, yeah, we, we get like the startup businesses but then as they mature, they kind of go somewhere else. And so I actually, I think we were talking with marketing about like using them for some, we should go back and make some cards.swyx: Yeah, yeah, yeah. You know, I remember, you know, as a very, very small breadth investor, I was like, why are we spending time like, doing these like stunts for GPUs? Like, you know, I think like as a, you know, typical like cloud hard hardware person, you go into an AWS you pick like T five X xl, whatever, and it's just like from a list and you look at the specs like, why animate this GP?And, and I, I do think like it just shows the level of care that goes throughout birth and Yeah. And now, and also the, and,Nader: and Nvidia. I think that's what the, the thing that struck me most when we first came in was like the amount of passion that everyone has. Like, I think, um, you know, you talk to, you talk to Kyle, you talk to, like, every VP that I've met at Nvidia goes so close to the metal.Like, I remember it was almost a year ago, and like my VP asked me, he's like, Hey, [00:07:00] what's cursor? And like, are you using it? And if so, why? Surprised at this, and he downloaded Cursor and he was asking me to help him like, use it. And I thought that was, uh, or like, just show him what he, you know, why we were using it.And so, the amount of care that I think everyone has and the passion, appreciate, passion and appreciation for the moment. Right. This is a very unique time. So it's really cool to see everyone really like, uh, appreciate that.swyx: Yeah.Acquisition and DevEx Shiftswyx: One thing I wanted to do before we move over to sort of like research topics and, uh, the, the stuff that Kyle's working on is just tell the story of the acquisition, right?Like, not many people have been, been through an acquisition with Nvidia. What's it like? Uh, what, yeah, just anything you'd like to say.Nader: It's a crazy experience. I think, uh, you know, we were the thing that was the most exciting for us was. Our goal was just to make it easier for developers.We wanted to find access to GPUs, make it easier to do that. And then all, oh, actually your question about launchable. So launchable was just make one click exper, like one click deploys for any software on top of the GPU. Mm-hmm. And so what we really liked about Nvidia was that it felt like we just got a lot more resources to do all of that.I think, uh, you [00:08:00] know, NVIDIA's goal is to make things as easy for developers as possible. So there was a really nice like synergy there. I think that, you know, when it comes to like an acquisition, I think the amount that the soul of the products align, I think is gonna be. Is going speak to the success of the acquisition.Yeah. And so it in many ways feels like we're home. This is a really great outcome for us. Like we you know, I love brev.nvidia.com. Like you should, you should use it's, it's theKyle: front page for GPUs.Nader: Yeah. Yeah. If you want GP views,Kyle: you go there, getswyx: it there, and it's like internally is growing very quickly.I, I don't remember You said some stats there.Nader: Yeah, yeah, yeah. It's, uh, I, I wish I had the exact numbers, but like internally, externally, it's been growing really quickly. We've been working with a bunch of partners with a bunch of different customers and ISVs, if you have a solution that you want someone that runs on the GPU and you want people to use it quickly, we can bundle it up, uh, in a launchable and make it a one click run.If you're doing things and you want just like a sandbox or something to run on, right. Like open claw. Huge moment. Super exciting. Our, uh, and we'll talk into it more, but. You know, internally, people wanna run this, and you, we know we have to be really careful from the security implications. Do we let this run on the corporate network?Security's guidance was, Hey, [00:09:00] run this on breath, it's in, you know, it's, it's, it's a vm, it's sitting in the cloud, it's off the corporate network. It's isolated. And so that's been our stance internally and externally about how to even run something like open call while we figure out how to run these things securely.But yeah,swyx: I think there's also like, you almost like we're the right team at the right time when Nvidia is starting to invest a lot more in developer experience or whatever you call it. Yeah. Uh, UX or I don't know what you call it, like software. Like obviously NVIDIA is always invested in software, but like, there's like, this is like a different audience.Yeah. It's aNader: widerKyle: developer base.swyx: Yeah. Right.Nader: Yeah. Yeah. You know, it's funny, it's like, it's not, uh,swyx: so like, what, what is it called internally? What, what is this that people should be aware that is going on there?Nader: Uh, what, like developer experienceswyx: or, yeah, yeah. Is it's called just developer experience or is there like a broader strategy hereNader: in Nvidia?Um, Nvidia always wants to make a good developer experience. The thing is and a lot of the technology is just really complicated. Like, it's not, it's uh, you know, I think, um. The thing that's been really growing or the AI's growing is having a huge moment, not [00:10:00] because like, let's say data scientists in 2018, were quiet then and are much louder now.The pie is com, right? There's a whole bunch of new audiences. My mom's wondering what she's doing. My sister's learned, like taught herself how to code. Like the, um, you know, I, I actually think just generally AI's a big equalizer and you're seeing a more like technologically literate society, I guess.Like everyone's, everyone's learning how to code. Uh, there isn't really an excuse for that. And so building a good UX means that you really understand who your end user is. And when your end user becomes such a wide, uh, variety of people, then you have to almost like reinvent the practice, right? Yeah. You haveKyle: to, and actually build more developer ux, right?Because the, there are tiers of developer base that were added. You know, the, the hackers that are building on top of open claw, right? For example, have never used gpu. They don't know what kuda is. They, they, they just want to run something.Nader: Yeah.Kyle: You need new UX that is not just. Hey, you know, how do you program something in Cuda and run it?And then, and then we built, you know, like when Deep Learning was getting big, we built, we built Torch and, and, but so recently the amount of like [00:11:00] layers that are added to that developer stack has just exploded because AI has become ubiquitous. Everyone's using it in different ways. Yeah. It'sNader: moving fast in every direction.Vertical, horizontal.Vibhu: Yeah. You guys, you even take it down to hardware, like the DGX Spark, you know, it's, it's basically the same system as just throwing it up on big GPU cluster.Nader: Yeah, yeah, yeah. It's amazing. Blackwell.swyx: Yeah. Uh, we saw the preview at the last year's GTC and that was one of the better performing, uh, videos so far, and video coverage so far.Awesome. This will beat it. Um,Nader: that wasswyx: actually, we have fingersNader: crossed. Yeah.DGX Spark and Remote AccessNader: Even when Grace Blackwell or when, um, uh, DGX Spark was first coming out getting to be involved in that from the beginning of the developer experience. And it just comes back to what youswyx: were involved.Nader: Yeah. St. St.swyx: Mars.Nader: Yeah. Yeah. I mean from, it was just like, I, I got an email, we just got thrown into the loop and suddenly yeah, I, it was actually really funny ‘cause I'm still pretty fresh from the acquisition and I'm, I'm getting an email from a bunch of the engineering VPs about like, the new hardware, GPU chip, like we're, or not chip, but just GPU system that we're putting out.And I'm like, okay, cool. Matters. Now involved with this for the ux, I'm like. What am I gonna do [00:12:00] here? So, I remember the first meeting, I was just like kind of quiet as I was hearing engineering VPs talk about what this box could be, what it could do, how we should use it. And I remember, uh, one of the first ideas that people were idea was like, oh, the first thing that it was like, I think a quote was like, the first thing someone's gonna wanna do with this is get two of them and run a Kubernetes cluster on top of them.And I was like, oh, I think I know why I'm here. I was like, the first thing we're doing is easy. SSH into the machine. And then, and you know, just kind of like scoping it down of like, once you can do that every, you, like the person who wants to run a Kubernetes cluster onto Sparks has a higher propensity for pain, then, then you know someone who buys it and wants to run open Claw right now, right?If you can make sure that that's as effortless as possible, then the rest becomes easy. So there's a tool called Nvidia Sync. It just makes the SSH connection really simple. So, you know, if you think about it like. If you have a Mac, uh, or a PC or whatever, if you have a laptop and you buy this GPU and you want to use it, you should be able to use it like it's A-A-G-P-U in the cloud, right?Um, but there's all this friction of like, how do you actually get into that? That's part of [00:13:00] Revs value proposition is just, you know, there's a CLI that wraps SSH and makes it simple. And so our goal is just get you into that machine really easily. And one thing we just launched at CES, it's in, it's still in like early access.We're ironing out some kinks, but it should be ready by GTC. You can register your spark on Brev. And so now if youswyx: like remote managed yeah, local hardware. Single pane of glass. Yeah. Yeah. Because Brev can already manage other clouds anyway, right?Vibhu: Yeah, yeah. And you use the spark on Brev as well, right?Nader: Yeah. But yeah, exactly. So, so you, you, so you, you set it up at home you can run the command on it, and then it gets it's essentially it'll appear in your Brev account, and then you can take your laptop to a Starbucks or to a cafe, and you'll continue to use your, you can continue use your spark just like any other cloud node on Brev.Yeah. Yeah. And it's just like a pre-provisioned centerswyx: in yourNader: home. Yeah, exactly.swyx: Yeah. Yeah.Vibhu: Tiny little data center.Nader: Tiny little, the size ofVibhu: your phone.SOL Culture and Dynamo Setupswyx: One more thing before we move on to Kyle. Just have so many Jensen stories and I just love, love mining Jensen stories. Uh, my favorite so far is SOL. Uh, what is, yeah, what is S-O-L-S-O-LNader: is actually, i, I think [00:14:00] of all the lessons I've learned, that one's definitely my favorite.Kyle: It'll always stick with you.Nader: Yeah. Yeah. I, you know, in your startup, everything's existential, right? Like we've, we've run out of money. We were like, on the risk of, of losing payroll, we've had to contract our team because we l ran outta money. And so like, um, because of that you're really always forcing yourself to I to like understand the root cause of everything.If you get a date, if you get a timeline, you know exactly why that date or timeline is there. You're, you're pushing every boundary and like, you're not just say, you're not just accepting like a, a no. Just because. And so as you start to introduce more layers, as you start to become a much larger organization, SOL is is essentially like what is the physics, right?The speed of light moves at a certain speed. So if flight's moving some slower, then you know something's in the way. So before trying to like layer reality back in of like, why can't this be delivered at some date? Let's just understand the physics. What is the theoretical limit to like, uh, how fast this can go?And then start to tell me why. ‘cause otherwise people will start telling you why something can't be done. But actually I think any great leader's goal is just to create urgency. Yeah. [00:15:00] There's an infiniteKyle: create compelling events, right?Nader: Yeah.Kyle: Yeah. So l is a term video is used to instigate a compelling event.You say this is done. How do we get there? What is the minimum? As much as necessary, as little as possible thing that it takes for us to get exactly here and. It helps you just break through a bunch of noise.swyx: Yeah.Kyle: Instantly.swyx: One thing I'm unclear about is, can only Jensen use the SOL card? Like, oh, no, no, no.Not everyone get the b******t out because obviously it's Jensen, but like, can someone else be like, no, likeKyle: frontline engineers use it.Nader: Yeah. Every, I think it's not so much about like, get the b******t out. It's like, it's like, give me the root understanding, right? Like, if you tell me something takes three weeks, it like, well, what's the first principles?Yeah, the first principles. It's like, what's the, what? Like why is it three weeks? What is the actual yeah. What's the actual limit of why this is gonna take three weeks? If you're gonna, if you, if let's say you wanted to buy a new computer and someone told you it's gonna be here in five days, what's the SOL?Well, like the SOL is like, I could walk into a Best Buy and pick it up for you. Right? So then anything that's like beyond that is, and is that practical? Is that how we're gonna, you know, let's say give everyone in the [00:16:00] company a laptop, like obviously not. So then like that's the SOL and then it's like, okay, well if we have to get more than 10, suddenly there might be some, right?And so now we can kind of piece the reality back.swyx: So, so this is the. Paul Graham do things that don't scale. Yeah. And this is also the, what people would now call behi agency. Yeah.Kyle: It's actually really interesting because there's a, there's a second hardware angle to SOL that like doesn't come up for all the org sol is used like culturally at aswyx: media for everything.I'm also mining for like, I think that can be annoying sometimes. And like someone keeps going IOO you and you're like, guys, like we have to be stable. We have to, we to f*****g plan. Yeah.Kyle: It's an interesting balance.Nader: Yeah. I encounter that with like, actually just with, with Alec, right? ‘cause we, we have a new conference so we need to launch, we have, we have goals of what we wanna launch by, uh, by the conference and like, yeah.At the end of the day, where isswyx: this GTC?Nader: Um, well this is like, so we, I mean we did it for CES, we did for GT CDC before that we're doing it for GTC San Jose. So I mean, like every, you know, we have a new moment. Um, and we want to launch something. Yeah. And we want to do so at SOL and that does mean that some, there's some level of prioritization that needs [00:17:00] to happen.And so it, it is difficult, right? I think, um, you have to be careful with what you're pushing. You know, stability is important and that should be factored into S-O-L-S-O-L isn't just like, build everything and let it break, you know, that, that's part of the conversation. So as you're laying, layering in all the details, one of them might be, Hey, we could build this, but then it's not gonna be stable for X, y, z reasons.And so that was like, one of our conversations for CES was, you know, hey, like we, we can get this into early access registering your spark with brev. But there are a lot of things that we need to do in order to feel really comfortable from a security perspective, right? There's a lot of networking involved before we deliver that to users.So it's like, okay. Let's get this to a point where we can at least let people experiment with it. We had it in a booth, we had it in Jensen's keynote, and then let's go iron out all the networking kinks. And that's not easy. And so, uh, that can come later. And so that was the way that we layered that back in.Yeah. ButKyle: It's not really about saying like, you don't have to do the, the maintenance or operational work. It's more about saying, you know, it's kind of like [00:18:00] highlights how progress is incremental, right? Like, what is the minimum thing that we can get to. And then there's SOL for like every component after that.But there's the SOL to get you, get you to the, the starting line. And that, that's usually how it's asked. Yeah. On the other side, you know, like SOL came out of like hardware at Nvidia. Right. So SOL is like literally if we ran the accelerator or the GPU with like at basically full speed with like no other constraints, like how FAST would be able to make a program go.swyx: Yeah. Yeah. Right.Kyle: Soswyx: in, in training that like, you know, then you work back to like some percentage of like MFU for example.Kyle: Yeah, that's a, that's a great example. So like, there's an, there's an S-O-L-M-F-U, and then there's like, you know, what's practically achievable.swyx: Cool. Should we move on to sort of, uh, Kyle's side?Uh, Kyle, you're coming more from the data science world. And, uh, I, I mean I always, whenever, whenever I meet someone who's done working in tabular stuff, graph neural networks, time series, these are basically when I go to new reps, I go to ICML, I walk the back halls. There's always like a small group of graph people.Yes. Absolute small group of tabular people. [00:19:00] And like, there's no one there. And like, it's very like, you know what I mean? Like, yeah, no, like it's, it's important interesting work if you care about solving the problems that they solve.Kyle: Yeah.swyx: But everyone else is just LMS all the time.Kyle: Yeah. I mean it's like, it's like the black hole, right?Has the event horizon reached this yet in nerves? Um,swyx: but like, you know, those are, those are transformers too. Yeah. And, and those are also like interesting things. Anyway, uh, I just wanted to spend a little bit of time on, on those, that background before we go into Dynamo, uh, proper.Kyle: Yeah, sure. I took a different path to Nvidia than that, or I joined six years ago, seven, if you count, when I was an intern.So I joined Nvidia, like right outta college. And the first thing I jumped into was not what I'd done in, during internship, which was like, you know, like some stuff for autonomous vehicles, like heavyweight object detection. I jumped into like, you know, something, I'm like, recommenders, this is popular. Andswyx: yeah, he did RexiKyle: as well.Yeah, Rexi. Yeah. I mean that, that was the taboo data at the time, right? You have tables of like, audience qualities and item qualities, and you're trying to figure out like which member of [00:20:00] the audience matches which item or, or more practically which item matches which member of the audience. And at the time, really it was like we were trying to enable.Uh, recommender, which had historically been like a little bit of a CP based workflow into something that like, ran really well in GPUs. And it's since been done. Like there are a bunch of libraries for Axis that run on GPUs. Uh, the common models like Deeplearning recommendation model, which came outta meta and the wide and deep model, which was used or was released by Google were very accelerated by GPUs using, you know, the fast HBM on the chips, especially to do, you know, vector lookups.But it was very interesting at the time and super, super relevant because like we were starting to get like. This explosion of feeds and things that required rec recommenders to just actively be on all the time. And sort of transitioned that a little bit towards graph neural networks when I discovered them because I was like, okay, you can actually use graphical neural networks to represent like, relationships between people, items, concepts, and that, that interested me.So I jumped into that at [00:21:00] Nvidia and, and got really involved for like two-ish years.swyx: Yeah. Uh, and something I learned from Brian Zaro Yeah. Is that you can just kind of choose your own path in Nvidia.Kyle: Oh my God. Yeah.swyx: Which is not a normal big Corp thing. Yeah. Like you, you have a lane, you stay in your lane.Nader: I think probably the reason why I enjoy being in a, a big company, the mission is the boss probably from a startup guy. Yeah. The missionswyx: is the boss.Nader: Yeah. Uh, it feels like a big game of pickup basketball. Like, you know, if you play one, if you wanna play basketball, you just go up to the court and you're like, Hey look, we're gonna play this game and we need three.Yeah. And you just like find your three. That's honestly for every new initiative that's what it feels like. Yeah.Vibhu: It also like shows, right? Like Nvidia. Just releasing state-of-the-art stuff in every domain. Yeah. Like, okay, you expect foundation models with Nemo tron voice just randomly parakeet.Call parakeet just comes out another one, uh, voice. TheKyle: video voice team has always been producing.Vibhu: Yeah. There's always just every other domain of paper that comes out, dataset that comes out. It's like, I mean, it also stems back to what Nvidia has to do, right? You have to make chips years before they're actually produced.Right? So you need to know, you need to really [00:22:00] focus. TheKyle: design process starts likeVibhu: exactlyKyle: three to five years before the chip gets to the market.Vibhu: Yeah. I, I'm curious more about what that's like, right? So like, you have specialist teams. Is it just like, you know, people find an interest, you go in, you go deep on whatever, and that kind of feeds back into, you know, okay, we, we expect predictions.Like the internals at Nvidia must be crazy. Right? You know? Yeah. Yeah. You know, you, you must. Not even without selling to people, you have your own predictions of where things are going. Yeah. And they're very based, very grounded. Right?Kyle: Yeah. It, it, it's really interesting. So there's like two things that I think that Amed does, which are quite interesting.Uh, one is like, we really index into passion. There's a big. Sort of organizational top sound push to like ensure that people are working on the things that they're passionate about. So if someone proposes something that's interesting, many times they can just email someone like way up the chain that they would find this relevant and say like, Hey, can I go work on this?Nader: It's actually like I worked at a, a big company for a couple years before, uh, starting on my startup journey and like, it felt very weird if you were to like email out of chain, if that makes [00:23:00] sense. Yeah. The emails at Nvidia are like mosh pitsswyx: shoot,Nader: and it's just like 60 people, just whatever. And like they're, there's this,swyx: they got messy like, reply all you,Nader: oh, it's in, it's insane.It's insane. They justKyle: help. You know, Maxim,Nader: the context. But, but that's actually like, I've actually, so this is a weird thing where I used to be like, why would we send emails? We have Slack. I am the entire, I'm the exact opposite. I feel so bad for anyone who's like messaging me on Slack ‘cause I'm so unresponsive.swyx: Your emailNader: Maxi, email Maxim. I'm email maxing Now email is a different, email is perfect because man, we can't work together. I'm email is great, right? Because important threads get bumped back up, right? Yeah, yeah. Um, and so Slack doesn't do that. So I just have like this casino going off on the right or on the left and like, I don't know which thread was from where or what, but like the threads get And then also just like the subject, so you can have like working threads.I think what's difficult is like when you're small, if you're just not 40,000 people I think Slack will work fine, but there's, I don't know what the inflection point is. There is gonna be a point where that becomes really messy and you'll actually prefer having email. ‘cause you can have working threads.You can cc more than nine people in a thread.Kyle: You can fork stuff.Nader: You can [00:24:00] fork stuff, which is super nice and just like y Yeah. And so, but that is part of where you can propose a plan. You can also just. Start, honestly, momentum's the only authority, right? So like, if you can just start, start to make a little bit of progress and show someone something, and then they can try it.That's, I think what's been, you know, I think the most effective way to push anything for forward. And that's both at Nvidia and I think just generally.Kyle: Yeah, there's, there's the other concept that like is explored a lot at Nvidia, which is this idea of a zero billion dollar business. Like market creation is a big thing at Nvidia.Like,swyx: oh, you want to go and start a zero billion dollar business?Kyle: Jensen says, we are completely happy investing in zero billion dollar markets. We don't care if this creates revenue. It's important for us to know about this market. We think it will be important in the future. It can be zero billion dollars for a while.I'm probably minging as words here for, but like, you know, like, I'll give an example. NVIDIA's been working on autonomous driving for a a long time,swyx: like an Nvidia car.Kyle: No, they, they'veVibhu: used the Mercedes, right? They're around the HQ and I think it finally just got licensed out. Now they're starting to be used quite a [00:25:00] bit.For 10 years you've been seeing Mercedes with Nvidia logos driving.Kyle: If you're in like the South San Santa Clara, it's, it's actually from South. Yeah. So, um. Zero billion dollar markets are, are a thing like, you know, Jensen,swyx: I mean, okay, look, cars are not a zero billion dollar market. But yeah, that's a bad example.Nader: I think, I think he's, he's messaging, uh, zero today, but, or even like internally, right? Like, like it's like, uh, an org doesn't have to ruthlessly find revenue very quickly to justify their existence. Right. Like a lot of the important research, a lot of the important technology being developed that, that's kind ofKyle: where research, research is very ide ideologically free at Nvidia.Yeah. Like they can pursue things that they wereswyx: Were you research officially?Kyle: I was never in research. Officially. I was always in engineering. Yeah. We in, I'm in an org called Deep Warning Algorithms, which is basically just how do we make things that are relevant to deep warning go fast.swyx: That sounds freaking cool.Vibhu: And I think a lot of that is underappreciated, right? Like time series. This week Google put out time. FF paper. Yeah. A new time series, paper res. Uh, Symantec, ID [00:26:00] started applying Transformers LMS to Yes. Rec system. Yes. And when you think the scale of companies deploying these right. Amazon recommendations, Google web search, it's like, it's huge scale andKyle: Yeah.Vibhu: You want fast?Kyle: Yeah. Yeah. Yeah. Actually it's, it, I, there's a fun moment that brought me like full circle. Like, uh, Amazon Ads recently gave a talk where they talked about using Dynamo for generative recommendation, which was like super, like weirdly cathartic for me. I'm like, oh my God. I've, I've supplanted what I was working on.Like, I, you're using LMS now to do what I was doing five years ago.swyx: Yeah. Amazing. And let's go right into Dynamo. Uh, maybe introduce Yeah, sure. To the top down and Yeah.Kyle: I think at this point a lot of people are familiar with the term of inference. Like funnily enough, like I went from, you know, inference being like a really niche topic to being something that's like discussed on like normal people's Twitter feeds.It's,Nader: it's on billboardsKyle: here now. Yeah. Very, very strange. Driving, driving, seeing just an inference ad on 1 0 1 inference at scale is becoming a lot more important. Uh, we have these moments like, you know, open claw where you have these [00:27:00] agents that take lots and lots of tokens, but produce, incredible results.There are many different aspects of test time scaling so that, you know, you can use more inference to generate a better result than if you were to use like a short amount of inference. There's reasoning, there's quiring, there's, adding agency to the model, allowing it to call tools and use skills.Dyno sort came about at Nvidia. Because myself and a couple others were, were sort of talking about the, these concepts that like, you know, you have inference engines like VLMS, shelan, tenor, TLM and they have like one single copy. They, they, they sort of think about like things as like one single copy, like one replica, right?Why Scale Out WinsKyle: Like one version of the model. But when you're actually serving things at scale, you can't just scale up that replica because you end up with like performance problems. There's a scaling limit to scaling up replicas. So you actually have to scale out to use a, maybe some Kubernetes type terminology.We kind of realized that there was like. A lot of potential optimization that we could do in scaling out and building systems for data [00:28:00] center scale inference. So Dynamo is this data center scale inference engine that sits on top of the frameworks like VLM Shilling and 10 T lm and just makes things go faster because you can leverage the economy of scale.The fact that you have KV cash, which we can define a little bit later, uh, in all these machines that is like unique and you wanna figure out like the ways to maximize your cash hits or you want to employ new techniques in inference like disaggregation, which Dynamo had introduced to the world in, in, in March, not introduced, it was a academic talk, but beforehand.But we are, you know, one of the first frameworks to start, supporting it. And we wanna like, sort of combine all these techniques into sort of a modular framework that allows you to. Accelerate your inference at scale.Nader: By the way, Kyle and I became friends on my first date, Nvidia, and I always loved, ‘cause like he always teaches meswyx: new things.Yeah. By the way, this is why I wanted to put two of you together. I was like, yeah, this is, this is gonna beKyle: good. It's very, it's very different, you know, like we've, we, we've, we've talked to each other a bunch [00:29:00] actually, you asked like, why, why can't we scale up?Nader: Yeah.Scale Up Limits ExplainedNader: model, you said model replicas.Kyle: Yeah. So you, so scale up means assigning moreswyx: heavier?Kyle: Yeah, heavier. Like making things heavier. Yeah, adding more GPUs. Adding more CPUs. Scale out is just like having a barrier saying, I'm gonna duplicate my representation of the model or a representation of this microservice or something, and I'm gonna like, replicate it Many times.Handle, load. And the reason that you can't scale, scale up, uh, past some points is like, you know, there, there, there are sort of hardware bounds and algorithmic bounds on, on that type of scaling. So I'll give you a good example that's like very trivial. Let's say you're on an H 100. The Maxim ENV link domain for H 100, for most Ds H one hundreds is heus, right?So if you scaled up past that, you're gonna have to figure out ways to handle the fact that now for the GPUs to communicate, you have to do it over Infin band, which is still very fast, but is not as fast as ENV link.swyx: Is it like one order of magnitude, like hundreds or,Kyle: it's about an order of magnitude?Yeah. Okay. Um, soswyx: not terrible.Kyle: [00:30:00] Yeah. I, I need to, I need to remember the, the data sheet here, like, I think it's like about 500 gigabytes. Uh, a second unidirectional for ENV link, and about 50 gigabytes a second unidirectional for Infin Band. I, it, it depends on the, the generation.swyx: I just wanna set this up for people who are not familiar with these kinds of like layers and the trash speedVibhu: and all that.Of course.From Laptop to Multi NodeVibhu: Also, maybe even just going like a few steps back before that, like most people are very familiar with. You see a, you know, you can use on your laptop, whatever these steel viol, lm you can just run inference there. All, there's all, you can, youcan run it on thatVibhu: laptop. You can run on laptop.Then you get to, okay, uh, models got pretty big, right? JLM five, they doubled the size, so mm-hmm. Uh, what do you do when you have to go from, okay, I can get 128 gigs of memory. I can run it on a spark. Then you have to go multi GPU. Yeah. Okay. Multi GPU, there's some support there. Now, if I'm a company and I don't have like.I'm not hiring the best researchers for this. Right. But I need to go [00:31:00] multi-node, right? I have a lot of servers. Okay, now there's efficiency problems, right? You can have multiple eight H 100 nodes, but, you know, is that as a, like, how do you do that efficiently?Kyle: Yeah. How do you like represent them? How do you choose how to represent the model?Yeah, exactly right. That's a, that's like a hard question. Everyone asks, how do you size oh, I wanna run GLM five, which just came out new model. There have been like four of them in the past week, by the way, like a bunch of new models.swyx: You know why? Right? Deep seek.Kyle: No comment. Oh. Yeah, but Ggl, LM five, right?We, we have this, new model. It's, it's like a large size, and you have to figure out how to both scale up and scale out, right? Because you have to find the right representation that you care about. Everyone does this differently. Let's be very clear. Everyone figures this out in their own path.Nader: I feel like a lot of AI or ML even is like, is like this. I think people think, you know, I, I was, there was some tweet a few months ago that was like, why hasn't fine tuning as a service taken off? You know, that might be me. It might have been you. Yeah. But people want it to be such an easy recipe to follow.But even like if you look at an ML model and specificKyle: to you Yeah,Nader: yeah.Kyle: And the [00:32:00] model,Nader: the situation, and there's just so much tinkering, right? Like when you see a model that has however many experts in the ME model, it's like, why that many experts? I don't, they, you know, they tried a bunch of things and that one seemed to do better.I think when it comes to how you're serving inference, you know, you have a bunch of decisions to make and there you can always argue that you can take something and make it more optimal. But I think it's this internal calibration and appetite for continued calibration.Vibhu: Yeah. And that doesn't mean like, you know, people aren't taking a shot at this, like tinker from thinking machines, you know?Yeah. RL as a service. Yeah, totally. It's, it also gets even harder when you try to do big model training, right? We're not the best at training Moes, uh, when they're pre-trained. Like we saw this with LAMA three, right? They're trained in such a sparse way that meta knows there's gonna be a bunch of inference done on these, right?They'll open source it, but it's very trained for what meta infrastructure wants, right? They wanna, they wanna inference it a lot. Now the question to basically think about is, okay, say you wanna serve a chat application, a coding copilot, right? You're doing a layer of rl, you're serving a model for X amount of people.Is it a chat model, a coding model? Dynamo, you know, back to that,Kyle: it's [00:33:00] like, yeah, sorry. So you we, we sort of like jumped off of, you know, jumped, uh, on that topic. Everyone has like, their own, own journey.Cost Quality Latency TradeoffsKyle: And I, I like to think of it as defined by like, what is the model you need? What is the accuracy you need?Actually I talked to NA about this earlier. There's three axes you care about. What is the quality that you're able to produce? So like, are you accurate enough or can you complete the task with enough, performance, high enough performance. Yeah, yeah. Uh, there's cost. Can you serve the model or serve your workflow?Because it's not just the model anymore, it's the workflow. It's the multi turn with an agent cheaply enough. And then can you serve it fast enough? And we're seeing all three of these, like, play out, like we saw, we saw new models from OpenAI that you know, are faster. You have like these new fast versions of models.You can change the amount of thinking to change the amount of quality, right? Produce more tokens, but at a higher cost in a, in a higher latency. And really like when you start this journey of like trying to figure out how you wanna host a model, you, you, you think about three things. What is the model I need to serve?How many times do I need to call it? What is the input sequence link was [00:34:00] the, what does the workflow look like on top of it? What is the SLA, what is the latency SLA that I need to achieve? Because there's usually some, this is usually like a constant, you, you know, the SLA that you need to hit and then like you try and find the lowest cost version that hits all of these constraints.Usually, you know, you, you start with those things and you say you, you kind of do like a bit of experimentation across some common configurations. You change the tensor parallel size, which is a form of parallelismVibhu: I take, it goes even deeper first. Gotta think what model.Kyle: Yes, course,ofKyle: course. It's like, it's like a multi-step design process because as you said, you can, you can choose a smaller model and then do more test time scaling and it'll equate the quality of a larger model because you're doing the test time scaling or you're adding a harness or something.So yes, it, it goes way deeper than that. But from the performance perspective, like once you get to the model you need, you need to host, you look at that and you say, Hey. I have this model, I need to serve it at the speed. What is the right configuration for that?Nader: You guys see the recent, uh, there was a paper I just saw like a few days ago that, uh, if you run [00:35:00] the same prompt twice, you're getting like double Just try itagain.Nader: Yeah, exactly.Vibhu: And you get a lot. Yeah. But the, the key thing there is you give the context of the failed try, right? Yeah. So it takes a shot. And this has been like, you know, basic guidance for quite a while. Just try again. ‘cause you know, trying, just try again. Did you try again? All adviceNader: in life.Vibhu: Just, it's a paper from Google, if I'm not mistaken, right?Yeah,Vibhu: yeah. I think it, it's like a seven bas little short paper. Yeah. Yeah. The title's very cute. And it's just like, yeah, just try again. Give it ask context,Kyle: multi-shot. You just like, say like, hey, like, you know, like take, take a little bit more, take a little bit more information, try and fail. Fail.Vibhu: And that basic concept has gone pretty deep.There's like, um, self distillation, rl where you, you do self distillation, you do rl and you have past failure and you know, that gives some signal so people take, try it again. Not strong enough.swyx: Uh, for, for listeners, uh, who listen to here, uh, vivo actually, and I, and we run a second YouTube channel for our paper club where, oh, that's awesome.Vivo just covered this. Yeah. Awesome. Self desolation and all that's, that's why he, to speed [00:36:00] on it.Nader: I'll to check it out.swyx: Yeah. It, it's just a good practice, like everyone needs, like a paper club where like you just read papers together and the social pressure just kind of forces you to just,Nader: we, we,there'sNader: like a big inference.Kyle: ReadingNader: group at a video. I feel so bad every time. I I, he put it on like, on our, he shared it.swyx: One, one ofNader: your guys,swyx: uh, is, is big in that, I forget es han Yeah, yeah,Kyle: es Han's on my team. Actually. Funny. There's a, there's a, there's a employee transfer between us. Han worked for Nater at Brev, and now he, he's on my team.He wasNader: our head of ai. And then, yeah, once we got in, andswyx: because I'm always looking for like, okay, can, can I start at another podcast that only does that thing? Yeah. And, uh, Esan was like, I was trying to like nudge Esan into like, is there something here? I mean, I don't think there's, there's new infant techniques every day.So it's like, it's likeKyle: you would, you would actually be surprised, um, the amount of blog posts you see. And ifswyx: there's a period where it was like, Medusa hydra, what Eagle, like, youKyle: know, now we have new forms of decode, uh, we have new forms of specula, of decoding or new,swyx: what,Kyle: what are youVibhu: excited? And it's exciting when you guys put out something like Tron.‘cause I remember the paper on this Tron three, [00:37:00] uh, the amount of like post train, the on tokens that the GPU rich can just train on. And it, it was a hybrid state space model, right? Yeah.Kyle: It's co-designed for the hardware.Vibhu: Yeah, go design for the hardware. And one of the things was always, you know, the state space models don't scale as well when you do a conversion or whatever the performance.And you guys are like, no, just keep draining. And Nitron shows a lot of that. Yeah.Nader: Also, something cool about Nitron it was released in layers, if you will, very similar to Dynamo. It's, it's, it's essentially it was released as you can, the pre-training, post-training data sets are released. Yeah. The recipes on how to do it are released.The model itself is released. It's full model. You just benefit from us turning on the GPUs. But there are companies like, uh, ServiceNow took the dataset and they trained their own model and we were super excited and like, you know, celebrated that work.ZoomVibhu: different. Zoom is, zoom is CGI, I think, uh, you know, also just to add like a lot of models don't put out based models and if there's that, why is fine tuning not taken off?You know, you can do your own training. Yeah,Kyle: sure.Vibhu: You guys put out based model, I think you put out everything.Nader: I believe I know [00:38:00]swyx: about base. BasicallyVibhu: without baseswyx: basic can be cancelable.Vibhu: Yeah. Base can be cancelable.swyx: Yeah.Vibhu: Safety training.swyx: Did we get a full picture of dymo? I, I don't know if we, what,Nader: what I'd love is you, you mentioned the three axes like break it down of like, you know, what's prefilled decode and like what are the optimizations that we can get with Dynamo?Kyle: Yeah. That, that's, that's, that's a great point. So to summarize on that three axis problem, right, there are three things that determine whether or not something can be done with inference, cost, quality, latency, right? Dynamo is supposed to be there to provide you like the runtime that allows you to pull levers to, you know, mix it up and move around the parade of frontier or the preto surface that determines is this actually possible with inference And AI todayNader: gives you the knobs.Kyle: Yeah, exactly. It gives you the knobs.Disaggregation Prefill vs DecodeKyle: Uh, and one thing that like we, we use a lot in contemporary inference and is, you know, starting to like pick up from, you know, in, in general knowledge is this co concept of disaggregation. So historically. Models would be hosted with a single inference engine. And that inference engine [00:39:00] would ping pong between two phases.There's prefill where you're reading the sequence generating KV cache, which is basically just a set of vectors that represent the sequence. And then using that KV cache to generate new tokens, which is called Decode. And some brilliant researchers across multiple different papers essentially made the realization that if you separate these two phases, you actually gain some benefits.Those benefits are basically a you don't have to worry about step synchronous scheduling. So the way that an inference engine works is you do one step and then you finish it, and then you schedule, you start scheduling the next step there. It's not like fully asynchronous. And the problem with that is you would have, uh, essentially pre-fill and decode are, are actually very different in terms of both their resource requirements and their sometimes their runtime.So you would have like prefill that would like block decode steps because you, you'd still be pre-filing and you couldn't schedule because you know the step has to end. So you remove that scheduling issue and then you also allow you, or you yourself, to like [00:40:00] split the work into two different ki types of pools.So pre-fill typically, and, and this changes as, as model architecture changes. Pre-fill is, right now, compute bound most of the time with the sequence is sufficiently long. It's compute bound. On the decode side because you're doing a full Passover, all the weights and the entire sequence, every time you do a decode step and you're, you don't have the quadratic computation of KV cache, it's usually memory bound because you're retrieving a linear amount of memory and you're doing a linear amount of compute as opposed to prefill where you retrieve a linear amount of memory and then use a quadratic.You know,Nader: it's funny, someone exo Labs did a really cool demo where for the DGX Spark, which has a lot more compute, you can do the pre the compute hungry prefill on a DG X spark and then do the decode on a, on a Mac. Yeah. And soVibhu: that's faster.Nader: Yeah. Yeah.Kyle: So you could, you can do that. You can do machine strat stratification.Nader: Yeah.Kyle: And like with our future generation generations of hardware, we actually announced, like with Reuben, this [00:41:00] new accelerator that is prefilled specific. It's called Reuben, CPX. SoKubernetes Scaling with GroveNader: I have a question when you do the scale out. Yeah. Is scaling out easier with Dynamo? Because when you need a new node, you can dedicate it to either the Prefill or, uh, decode.Kyle: Yeah. So Dynamo actually has like a, a Kubernetes component in it called Grove that allows you to, to do this like crazy scaling specialization. It has like this hot, it's a representation that, I don't wanna go too deep into Kubernetes here, but there was a previous way that you would like launch multi-node work.Uh, it's called Leader Worker Set. It's in the Kubernetes standard, and Leader worker set is great. It served a lot of people super well for a long period of time. But one of the things that it's struggles with is representing a set of cases where you have a multi-node replica that has a pair, right?You know, prefill and decode, or it's not paired, but it has like a second stage that has a ratio that changes over time. And prefill and decode are like two different things as your workload changes, right? The amount of prefill you'll need to do may change. [00:42:00] The amount of decode that you, you'll need to do might change, right?Like, let's say you start getting like insanely long queries, right? That probably means that your prefill scales like harder because you're hitting these, this quadratic scaling growth.swyx: Yeah.And then for listeners, like prefill will be long input. Decode would be long output, for example, right?Kyle: Yeah. So like decode, decode scale. I mean, decode is funny because the amount of tokens that you produce scales with the output length, but the amount of work that you do per step scales with the amount of tokens in the context.swyx: Yes.Kyle: So both scales with the input and the output.swyx: That's true.Kyle: But on the pre-fold view code side, like if.Suddenly, like the amount of work you're doing on the decode side stays about the same or like scales a little bit, and then the prefilled side like jumps up a lot. You actually don't want that ratio to be the same. You want it to change over time. So Dynamo has a set of components that A, tell you how to scale.It tells you how many prefilled workers and decoded workers you, it thinks you should have, and also provides a scheduling API for Kubernetes that allows you to actually represent and affect this scheduling on, on, on your actual [00:43:00] hardware, on your compute infrastructure.Nader: Not gonna lie. I feel a little embarrassed for being proud of my SVG function earlier.swyx: No, itNader: wasreallyKyle: cute. I, Iswyx: likeNader: it's all,swyx: it's all engineering. It's all engineering. Um, that's where I'mKyle: technical.swyx: One thing I'm, I'm kind of just curious about with all with you see at a systems level, everything going on here. Mm-hmm. And we, you know, we're scaling it up in, in multi, in distributed systems.Context Length and Co Designswyx: Um, I think one thing that's like kind of, of the moment right now is people are asking, is there any SOL sort of upper bounds. In terms of like, let's call, just call it context length for one for of a better word, but you can break it down however you like.Nader: Yeah.swyx: I just think like, well, yeah, I mean, like clearly you can engage in hybrid architectures and throw in some state space models in there.All, all you want, but it looks, still looks very attention heavy.Kyle: Yes. Uh, yeah. Long context is attention heavy. I mean, we have these hybrid models, um,swyx: to take and most, most models like cap out at a million contexts and that's it. Yeah. Like for the last two years has been it.Kyle: Yeah. The model hardware context co-design thing that we're seeing these days is actually super [00:44:00] interesting.It's like my, my passion, like my secret side passion. We see models like Kimmy or G-P-T-O-S-S. I'm use these because I, I know specific things about these models. So Kimmy two comes out, right? And it's an interesting model. It's like, like a deep seek style architecture is MLA. It's basically deep seek, scaled like a little bit differently, um, and obviously trained differently as well.But they, they talked about, why they made the design choices for context. Kimmy has more experts, but fewer attention heads, and I believe a slightly smaller attention, uh, like dimension. But I need to remember, I need to check that. Uh, it doesn't matter. But they discussed this actually at length in a blog post on ji, which is like our pu which is like credit puswyx: Yeah.Kyle: Um, in, in China. Chinese red.swyx: Yeah.Kyle: It's, yeah. So it, it's, it's actually an incredible blog post. Uh, like all the mls people in, in, in that, I've seen that on GPU are like very brilliant, but they, they talk about like the creators of Kimi K two [00:45:00] actually like, talked about it on, on, on there in the blog post.And they say, we, we actually did an experiment, right? Attention scales with the number of heads, obviously. Like if you have 64 heads versus 32 heads, you do half the work of attention. You still scale quadratic, but you do half the work. And they made a, a very specific like. Sort of barter in their system, in their architecture, they basically said, Hey, what if we gave it more experts, so we're gonna use more memory capacity.But we keep the amount of activated experts the same. We increase the expert sparsity, so we have fewer experts act. The ratio to of experts activated to number of experts is smaller, and we decrease the number of attention heads.Vibhu: And kind of for context, what the, what we had been seeing was you make models sparser instead.So no one was really touching heads. You're just having, uh,Kyle: well, they, they did, they implicitly made it sparser.Vibhu: Yeah, yeah. For, for Kimmy. They did,Kyle: yes.Vibhu: They also made it sparser. But basically what we were seeing was people were at the level of, okay, there's a sparsity ratio. You want more total parameters, less active, and that's sparsity.[00:46:00]But what you see from papers, like, the labs like moonshot deep seek, they go to the level of, okay, outside of just number of experts, you can also change how many attention heads and less attention layers. More attention. Layers. Layers, yeah. Yes, yes. So, and that's all basically coming back to, just tied together is like hardware model, co-design, which isKyle: hardware model, co model, context, co-design.Vibhu: Yeah.Kyle: Right. Like if you were training a, a model that was like. Really, really short context, uh, or like really is good at super short context tasks. You may like design it in a way such that like you don't care about attention scaling because it hasn't hit that, like the turning point where like the quadratic curve takes over.Nader: How do you consider attention or context as a separate part of the co-design? Like I would imagine hardware or just how I would've thought of it is like hardware model. Co-design would be hardware model context co-designKyle: because the harness and the context that is produced by the harness is a part of the model.Once it's trained in,Vibhu: like even though towards the end you'll do long context, you're not changing architecture through I see. Training. Yeah.Kyle: I mean you can try.swyx: You're saying [00:47:00] everyone's training the harness into the model.Kyle: I would say to some degree, orswyx: there's co-design for harness. I know there's a small amount, but I feel like not everyone has like gone full send on this.Kyle: I think, I think I think it's important to internalize the harness that you think the model will be running. Running into the model.swyx: Yeah. Interesting. Okay. Bash is like the universal harness,Kyle: right? Like I'll, I'll give. An example here, right? I mean, or just like a, like a, it's easy proof, right? If you can train against a harness and you're using that harness for everything, wouldn't you just train with the harness to ensure that you get the best possible quality out of,swyx: Well, the, uh, I, I can provide a counter argument.Yeah, sure. Which is what you wanna provide a generally useful model for other people to plug into their harnesses, right? So if youKyle: Yeah. Harnesses can be open, open source, right?swyx: Yeah. So I mean, that's, that's effectively what's happening with Codex.Kyle: Yeah.swyx: And, but like you may want like a different search tool and then you may have to name it differently or,Nader: I don't know how much people have pushed on this, but can you.Train a model, would it be, have you have people compared training a model for the for the harness versus [00:48:00] like post training forswyx: I think it's the same thing. It's the same thing. It's okay. Just extra post training. INader: see.swyx: And so, I mean, cognition does this course, it does this where you, you just have to like, if your tool is slightly different, um, either force your tool to be like the tool that they train for.Hmm. Or undo their training for their tool and then Oh, that's re retrain. Yeah. It's, it's really annoying and like,Kyle: I would hope that eventually we hit like a certain level of generality with respect to training newswyx: tools. This is not a GI like, it's, this is a really stupid like. Learn my tool b***h.Like, I don't know if, I don't know if I can say that, but like, you know, um, I think what my point kind of is, is that there's, like, I look at slopes of the scaling laws and like, this slope is not working, man. We, we are at a million token con

In this episode of The Canadian Investor Podcast, Simon Belanger and Dan Kent kick things off with a surprising ripple effect from the AI boom: a full-blown RAM/memory shortage that’s sending PC upgrade costs through the roof. They break down why high-bandwidth memory (HBM) is crowding out “normal” consumer RAM production, how Micron, Samsung, and SK Hynix are prioritizing the most profitable AI-driven demand, and what that could mean for pricing, upgrade cycles, and the broader tech supply chain. From there, they shift into a pragmatic, investor-focused look at positioning during geopolitical uncertainty—without cheerleading conflict. Dan outlines key areas investors often look at in these environments: defense contractors (and why buying after the headlines can be “buying the umbrella in the rain”), Canadian energy as a cleaner way to express higher oil prices with less Middle East exposure, the growing (and expensive) opportunity set in cybersecurity, and gold as both a safe haven and an inflation hedge. They also touch on different ways to gain exposure—individual names vs. ETFs—and wrap up with updates on the podcast’s YouTube live plans and what’s coming next. Tickers of Stocks discussed: LMT, NOC, GD, RTX, MU, AEM, FNV, WPM, ZJG.TO Subscribe to our Our New Youtube Channel! Check out our portfolio by going to Jointci.com Our Website Our New Youtube Channel! Canadian Investor Podcast Network Twitter: @cdn_investing Simon’s twitter: @Fiat_Iceberg Braden’s twitter: @BradoCapital Dan’s Twitter: @stocktrades_ca Want to learn more about Real Estate Investing? Check out the Canadian Real Estate Investor Podcast! Apple Podcast - The Canadian Real Estate Investor Spotify - The Canadian Real Estate Investor Web player - The Canadian Real Estate Investor Asset Allocation ETFs | BMO Global Asset Management Sign up for Fiscal.ai for free to get easy access to global stock coverage and powerful AI investing tools. Register for EQ Bank, the seamless digital banking experience with better rates and no nonsense. See omnystudio.com/listener for privacy information.

From rewriting Google's search stack in the early 2000s to reviving sparse trillion-parameter models and co-designing TPUs with frontier ML research, Jeff Dean has quietly shaped nearly every layer of the modern AI stack. As Chief AI Scientist at Google and a driving force behind Gemini, Jeff has lived through multiple scaling revolutions from CPUs and sharded indices to multimodal models that reason across text, video, and code.Jeff joins us to unpack what it really means to “own the Pareto frontier,” why distillation is the engine behind every Flash model breakthrough, how energy (in picojoules) not FLOPs is becoming the true bottleneck, what it was like leading the charge to unify all of Google's AI teams, and why the next leap won't come from bigger context windows alone, but from systems that give the illusion of attending to trillions of tokens.We discuss:* Jeff's early neural net thesis in 1990: parallel training before it was cool, why he believed scaling would win decades early, and the “bigger model, more data, better results” mantra that held for 15 years* The evolution of Google Search: sharding, moving the entire index into memory in 2001, softening query semantics pre-LLMs, and why retrieval pipelines already resemble modern LLM systems* Pareto frontier strategy: why you need both frontier “Pro” models and low-latency “Flash” models, and how distillation lets smaller models surpass prior generations* Distillation deep dive: ensembles → compression → logits as soft supervision, and why you need the biggest model to make the smallest one good* Latency as a first-class objective: why 10–50x lower latency changes UX entirely, and how future reasoning workloads will demand 10,000 tokens/sec* Energy-based thinking: picojoules per bit, why moving data costs 1000x more than a multiply, batching through the lens of energy, and speculative decoding as amortization* TPU co-design: predicting ML workloads 2–6 years out, speculative hardware features, precision reduction, sparsity, and the constant feedback loop between model architecture and silicon* Sparse models and “outrageously large” networks: trillions of parameters with 1–5% activation, and why sparsity was always the right abstraction* Unified vs. specialized models: abandoning symbolic systems, why general multimodal models tend to dominate vertical silos, and when vertical fine-tuning still makes sense* Long context and the illusion of scale: beyond needle-in-a-haystack benchmarks toward systems that narrow trillions of tokens to 117 relevant documents* Personalized AI: attending to your emails, photos, and documents (with permission), and why retrieval + reasoning will unlock deeply personal assistants* Coding agents: 50 AI interns, crisp specifications as a new core skill, and how ultra-low latency will reshape human–agent collaboration* Why ideas still matter: transformers, sparsity, RL, hardware, systems — scaling wasn't blind; the pieces had to multiply togetherShow Notes:* Gemma 3 Paper* Gemma 3* Gemini 2.5 Report* Jeff Dean's “Software Engineering Advice fromBuilding Large-Scale Distributed Systems” Presentation (with Back of the Envelope Calculations)* Latency Numbers Every Programmer Should Know by Jeff Dean* The Jeff Dean Facts* Jeff Dean Google Bio* Jeff Dean on “Important AI Trends” @Stanford AI Club* Jeff Dean & Noam Shazeer — 25 years at Google (Dwarkesh)—Jeff Dean* LinkedIn: https://www.linkedin.com/in/jeff-dean-8b212555* X: https://x.com/jeffdeanGoogle* https://google.com* https://deepmind.googleFull Video EpisodeTimestamps00:00:04 — Introduction: Alessio & Swyx welcome Jeff Dean, chief AI scientist at Google, to the Latent Space podcast00:00:30 — Owning the Pareto Frontier & balancing frontier vs low-latency models00:01:31 — Frontier models vs Flash models + role of distillation00:03:52 — History of distillation and its original motivation00:05:09 — Distillation's role in modern model scaling00:07:02 — Model hierarchy (Flash, Pro, Ultra) and distillation sources00:07:46 — Flash model economics & wide deployment00:08:10 — Latency importance for complex tasks00:09:19 — Saturation of some tasks and future frontier tasks00:11:26 — On benchmarks, public vs internal00:12:53 — Example long-context benchmarks & limitations00:15:01 — Long-context goals: attending to trillions of tokens00:16:26 — Realistic use cases beyond pure language00:18:04 — Multimodal reasoning and non-text modalities00:19:05 — Importance of vision & motion modalities00:20:11 — Video understanding example (extracting structured info)00:20:47 — Search ranking analogy for LLM retrieval00:23:08 — LLM representations vs keyword search00:24:06 — Early Google search evolution & in-memory index00:26:47 — Design principles for scalable systems00:28:55 — Real-time index updates & recrawl strategies00:30:06 — Classic “Latency numbers every programmer should know”00:32:09 — Cost of memory vs compute and energy emphasis00:34:33 — TPUs & hardware trade-offs for serving models00:35:57 — TPU design decisions & co-design with ML00:38:06 — Adapting model architecture to hardware00:39:50 — Alternatives: energy-based models, speculative decoding00:42:21 — Open research directions: complex workflows, RL00:44:56 — Non-verifiable RL domains & model evaluation00:46:13 — Transition away from symbolic systems toward unified LLMs00:47:59 — Unified models vs specialized ones00:50:38 — Knowledge vs reasoning & retrieval + reasoning00:52:24 — Vertical model specialization & modules00:55:21 — Token count considerations for vertical domains00:56:09 — Low resource languages & contextual learning00:59:22 — Origins: Dean's early neural network work01:10:07 — AI for coding & human–model interaction styles01:15:52 — Importance of crisp specification for coding agents01:19:23 — Prediction: personalized models & state retrieval01:22:36 — Token-per-second targets (10k+) and reasoning throughput01:23:20 — Episode conclusion and thanksTranscriptAlessio Fanelli [00:00:04]: Hey everyone, welcome to the Latent Space podcast. This is Alessio, founder of Kernel Labs, and I'm joined by Swyx, editor of Latent Space. Shawn Wang [00:00:11]: Hello, hello. We're here in the studio with Jeff Dean, chief AI scientist at Google. Welcome. Thanks for having me. It's a bit surreal to have you in the studio. I've watched so many of your talks, and obviously your career has been super legendary. So, I mean, congrats. I think the first thing must be said, congrats on owning the Pareto Frontier.Jeff Dean [00:00:30]: Thank you, thank you. Pareto Frontiers are good. It's good to be out there.Shawn Wang [00:00:34]: Yeah, I mean, I think it's a combination of both. You have to own the Pareto Frontier. You have to have like frontier capability, but also efficiency, and then offer that range of models that people like to use. And, you know, some part of this was started because of your hardware work. Some part of that is your model work, and I'm sure there's lots of secret sauce that you guys have worked on cumulatively. But, like, it's really impressive to see it all come together in, like, this slittily advanced.Jeff Dean [00:01:04]: Yeah, yeah. I mean, I think, as you say, it's not just one thing. It's like a whole bunch of things up and down the stack. And, you know, all of those really combine to help make UNOS able to make highly capable large models, as well as, you know, software techniques to get those large model capabilities into much smaller, lighter weight models that are, you know, much more cost effective and lower latency, but still, you know, quite capable for their size. Yeah.Alessio Fanelli [00:01:31]: How much pressure do you have on, like, having the lower bound of the Pareto Frontier, too? I think, like, the new labs are always trying to push the top performance frontier because they need to raise more money and all of that. And you guys have billions of users. And I think initially when you worked on the CPU, you were thinking about, you know, if everybody that used Google, we use the voice model for, like, three minutes a day, they were like, you need to double your CPU number. Like, what's that discussion today at Google? Like, how do you prioritize frontier versus, like, we have to do this? How do we actually need to deploy it if we build it?Jeff Dean [00:02:03]: Yeah, I mean, I think we always want to have models that are at the frontier or pushing the frontier because I think that's where you see what capabilities now exist that didn't exist at the sort of slightly less capable last year's version or last six months ago version. At the same time, you know, we know those are going to be really useful for a bunch of use cases, but they're going to be a bit slower and a bit more expensive than people might like for a bunch of other broader models. So I think what we want to do is always have kind of a highly capable sort of affordable model that enables a whole bunch of, you know, lower latency use cases. People can use them for agentic coding much more readily and then have the high-end, you know, frontier model that is really useful for, you know, deep reasoning, you know, solving really complicated math problems, those kinds of things. And it's not that. One or the other is useful. They're both useful. So I think we'd like to do both. And also, you know, through distillation, which is a key technique for making the smaller models more capable, you know, you have to have the frontier model in order to then distill it into your smaller model. So it's not like an either or choice. You sort of need that in order to actually get a highly capable, more modest size model. Yeah.Alessio Fanelli [00:03:24]: I mean, you and Jeffrey came up with the solution in 2014.Jeff Dean [00:03:28]: Don't forget, L'Oreal Vinyls as well. Yeah, yeah.Alessio Fanelli [00:03:30]: A long time ago. But like, I'm curious how you think about the cycle of these ideas, even like, you know, sparse models and, you know, how do you reevaluate them? How do you think about in the next generation of model, what is worth revisiting? Like, yeah, they're just kind of like, you know, you worked on so many ideas that end up being influential, but like in the moment, they might not feel that way necessarily. Yeah.Jeff Dean [00:03:52]: I mean, I think distillation was originally motivated because we were seeing that we had a very large image data set at the time, you know, 300 million images that we could train on. And we were seeing that if you create specialists for different subsets of those image categories, you know, this one's going to be really good at sort of mammals, and this one's going to be really good at sort of indoor room scenes or whatever, and you can cluster those categories and train on an enriched stream of data after you do pre-training on a much broader set of images. You get much better performance. If you then treat that whole set of maybe 50 models you've trained as a large ensemble, but that's not a very practical thing to serve, right? So distillation really came about from the idea of, okay, what if we want to actually serve that and train all these independent sort of expert models and then squish it into something that actually fits in a form factor that you can actually serve? And that's, you know, not that different from what we're doing today. You know, often today we're instead of having an ensemble of 50 models. We're having a much larger scale model that we then distill into a much smaller scale model.Shawn Wang [00:05:09]: Yeah. A part of me also wonders if distillation also has a story with the RL revolution. So let me maybe try to articulate what I mean by that, which is you can, RL basically spikes models in a certain part of the distribution. And then you have to sort of, well, you can spike models, but usually sometimes... It might be lossy in other areas and it's kind of like an uneven technique, but you can probably distill it back and you can, I think that the sort of general dream is to be able to advance capabilities without regressing on anything else. And I think like that, that whole capability merging without loss, I feel like it's like, you know, some part of that should be a distillation process, but I can't quite articulate it. I haven't seen much papers about it.Jeff Dean [00:06:01]: Yeah, I mean, I tend to think of one of the key advantages of distillation is that you can have a much smaller model and you can have a very large, you know, training data set and you can get utility out of making many passes over that data set because you're now getting the logits from the much larger model in order to sort of coax the right behavior out of the smaller model that you wouldn't otherwise get with just the hard labels. And so, you know, I think that's what we've observed. Is you can get, you know, very close to your largest model performance with distillation approaches. And that seems to be, you know, a nice sweet spot for a lot of people because it enables us to kind of, for multiple Gemini generations now, we've been able to make the sort of flash version of the next generation as good or even substantially better than the previous generations pro. And I think we're going to keep trying to do that because that seems like a good trend to follow.Shawn Wang [00:07:02]: So, Dara asked, so it was the original map was Flash Pro and Ultra. Are you just sitting on Ultra and distilling from that? Is that like the mother load?Jeff Dean [00:07:12]: I mean, we have a lot of different kinds of models. Some are internal ones that are not necessarily meant to be released or served. Some are, you know, our pro scale model and we can distill from that as well into our Flash scale model. So I think, you know, it's an important set of capabilities to have and also inference time scaling. It can also be a useful thing to improve the capabilities of the model.Shawn Wang [00:07:35]: And yeah, yeah, cool. Yeah. And obviously, I think the economy of Flash is what led to the total dominance. I think the latest number is like 50 trillion tokens. I don't know. I mean, obviously, it's changing every day.Jeff Dean [00:07:46]: Yeah, yeah. But, you know, by market share, hopefully up.Shawn Wang [00:07:50]: No, I mean, there's no I mean, there's just the economics wise, like because Flash is so economical, like you can use it for everything. Like it's in Gmail now. It's in YouTube. Like it's yeah. It's in everything.Jeff Dean [00:08:02]: We're using it more in our search products of various AI mode reviews.Shawn Wang [00:08:05]: Oh, my God. Flash past the AI mode. Oh, my God. Yeah, that's yeah, I didn't even think about that.Jeff Dean [00:08:10]: I mean, I think one of the things that is quite nice about the Flash model is not only is it more affordable, it's also a lower latency. And I think latency is actually a pretty important characteristic for these models because we're going to want models to do much more complicated things that are going to involve, you know, generating many more tokens from when you ask the model to do so. So, you know, if you're going to ask the model to do something until it actually finishes what you ask it to do, because you're going to ask now, not just write me a for loop, but like write me a whole software package to do X or Y or Z. And so having low latency systems that can do that seems really important. And Flash is one direction, one way of doing that. You know, obviously our hardware platforms enable a bunch of interesting aspects of our, you know, serving stack as well, like TPUs, the interconnect between. Chips on the TPUs is actually quite, quite high performance and quite amenable to, for example, long context kind of attention operations, you know, having sparse models with lots of experts. These kinds of things really, really matter a lot in terms of how do you make them servable at scale.Alessio Fanelli [00:09:19]: Yeah. Does it feel like there's some breaking point for like the proto Flash distillation, kind of like one generation delayed? I almost think about almost like the capability as a. In certain tasks, like the pro model today is a saturated, some sort of task. So next generation, that same task will be saturated at the Flash price point. And I think for most of the things that people use models for at some point, the Flash model in two generation will be able to do basically everything. And how do you make it economical to like keep pushing the pro frontier when a lot of the population will be okay with the Flash model? I'm curious how you think about that.Jeff Dean [00:09:59]: I mean, I think that's true. If your distribution of what people are asking people, the models to do is stationary, right? But I think what often happens is as the models become more capable, people ask them to do more, right? So, I mean, I think this happens in my own usage. Like I used to try our models a year ago for some sort of coding task, and it was okay at some simpler things, but wouldn't do work very well for more complicated things. And since then, we've improved dramatically on the more complicated coding tasks. And now I'll ask it to do much more complicated things. And I think that's true, not just of coding, but of, you know, now, you know, can you analyze all the, you know, renewable energy deployments in the world and give me a report on solar panel deployment or whatever. That's a very complicated, you know, more complicated task than people would have asked a year ago. And so you are going to want more capable models to push the frontier in the absence of what people ask the models to do. And that also then gives us. Insight into, okay, where does the, where do things break down? How can we improve the model in these, these particular areas, uh, in order to sort of, um, make the next generation even better.Alessio Fanelli [00:11:11]: Yeah. Are there any benchmarks or like test sets they use internally? Because it's almost like the same benchmarks get reported every time. And it's like, all right, it's like 99 instead of 97. Like, how do you have to keep pushing the team internally to it? Or like, this is what we're building towards. Yeah.Jeff Dean [00:11:26]: I mean, I think. Benchmarks, particularly external ones that are publicly available. Have their utility, but they often kind of have a lifespan of utility where they're introduced and maybe they're quite hard for current models. You know, I, I like to think of the best kinds of benchmarks are ones where the initial scores are like 10 to 20 or 30%, maybe, but not higher. And then you can sort of work on improving that capability for, uh, whatever it is, the benchmark is trying to assess and get it up to like 80, 90%, whatever. I, I think once it hits kind of 95% or something, you get very diminishing returns from really focusing on that benchmark, cuz it's sort of, it's either the case that you've now achieved that capability, or there's also the issue of leakage in public data or very related kind of data being, being in your training data. Um, so we have a bunch of held out internal benchmarks that we really look at where we know that wasn't represented in the training data at all. There are capabilities that we want the model to have. Um, yeah. Yeah. Um, that it doesn't have now, and then we can work on, you know, assessing, you know, how do we make the model better at these kinds of things? Is it, we need different kind of data to train on that's more specialized for this particular kind of task. Do we need, um, you know, a bunch of, uh, you know, architectural improvements or some sort of, uh, model capability improvements, you know, what would help make that better?Shawn Wang [00:12:53]: Is there, is there such an example that you, uh, a benchmark inspired in architectural improvement? Like, uh, I'm just kind of. Jumping on that because you just.Jeff Dean [00:13:02]: Uh, I mean, I think some of the long context capability of the, of the Gemini models that came, I guess, first in 1.5 really were about looking at, okay, we want to have, um, you know,Shawn Wang [00:13:15]: immediately everyone jumped to like completely green charts of like, everyone had, I was like, how did everyone crack this at the same time? Right. Yeah. Yeah.Jeff Dean [00:13:23]: I mean, I think, um, and once you're set, I mean, as you say that needed single needle and a half. Hey, stack benchmark is really saturated for at least context links up to 1, 2 and K or something. Don't actually have, you know, much larger than 1, 2 and 8 K these days or two or something. We're trying to push the frontier of 1 million or 2 million context, which is good because I think there are a lot of use cases where. Yeah. You know, putting a thousand pages of text or putting, you know, multiple hour long videos and the context and then actually being able to make use of that as useful. Try to, to explore the über graduation are fairly large. But the single needle in a haystack benchmark is sort of saturated. So you really want more complicated, sort of multi-needle or more realistic, take all this content and produce this kind of answer from a long context that sort of better assesses what it is people really want to do with long context. Which is not just, you know, can you tell me the product number for this particular thing?Shawn Wang [00:14:31]: Yeah, it's retrieval. It's retrieval within machine learning. It's interesting because I think the more meta level I'm trying to operate at here is you have a benchmark. You're like, okay, I see the architectural thing I need to do in order to go fix that. But should you do it? Because sometimes that's an inductive bias, basically. It's what Jason Wei, who used to work at Google, would say. Exactly the kind of thing. Yeah, you're going to win. Short term. Longer term, I don't know if that's going to scale. You might have to undo that.Jeff Dean [00:15:01]: I mean, I like to sort of not focus on exactly what solution we're going to derive, but what capability would you want? And I think we're very convinced that, you know, long context is useful, but it's way too short today. Right? Like, I think what you would really want is, can I attend to the internet while I answer my question? Right? But that's not going to happen. I think that's going to be solved by purely scaling the existing solutions, which are quadratic. So a million tokens kind of pushes what you can do. You're not going to do that to a trillion tokens, let alone, you know, a billion tokens, let alone a trillion. But I think if you could give the illusion that you can attend to trillions of tokens, that would be amazing. You'd find all kinds of uses for that. You would have attend to the internet. You could attend to the pixels of YouTube and the sort of deeper representations that we can find. You could attend to the form for a single video, but across many videos, you know, on a personal Gemini level, you could attend to all of your personal state with your permission. So like your emails, your photos, your docs, your plane tickets you have. I think that would be really, really useful. And the question is, how do you get algorithmic improvements and system level improvements that get you to something where you actually can attend to trillions of tokens? Right. In a meaningful way. Yeah.Shawn Wang [00:16:26]: But by the way, I think I did some math and it's like, if you spoke all day, every day for eight hours a day, you only generate a maximum of like a hundred K tokens, which like very comfortably fits.Jeff Dean [00:16:38]: Right. But if you then say, okay, I want to be able to understand everything people are putting on videos.Shawn Wang [00:16:46]: Well, also, I think that the classic example is you start going beyond language into like proteins and whatever else is extremely information dense. Yeah. Yeah.Jeff Dean [00:16:55]: I mean, I think one of the things about Gemini's multimodal aspects is we've always wanted it to be multimodal from the start. And so, you know, that sometimes to people means text and images and video sort of human-like and audio, audio, human-like modalities. But I think it's also really useful to have Gemini know about non-human modalities. Yeah. Like LIDAR sensor data from. Yes. Say, Waymo vehicles or. Like robots or, you know, various kinds of health modalities, x-rays and MRIs and imaging and genomics information. And I think there's probably hundreds of modalities of data where you'd like the model to be able to at least be exposed to the fact that this is an interesting modality and has certain meaning in the world. Where even if you haven't trained on all the LIDAR data or MRI data, you could have, because maybe that's not, you know, it doesn't make sense in terms of trade-offs of. You know, what you include in your main pre-training data mix, at least including a little bit of it is actually quite useful. Yeah. Because it sort of tempts the model that this is a thing.Shawn Wang [00:18:04]: Yeah. Do you believe, I mean, since we're on this topic and something I just get to ask you all the questions I always wanted to ask, which is fantastic. Like, are there some king modalities, like modalities that supersede all the other modalities? So a simple example was Vision can, on a pixel level, encode text. And DeepSeq had this DeepSeq CR paper that did that. Vision. And Vision has also been shown to maybe incorporate audio because you can do audio spectrograms and that's, that's also like a Vision capable thing. Like, so, so maybe Vision is just the king modality and like. Yeah.Jeff Dean [00:18:36]: I mean, Vision and Motion are quite important things, right? Motion. Well, like video as opposed to static images, because I mean, there's a reason evolution has evolved eyes like 23 independent ways, because it's such a useful capability for sensing the world around you, which is really what we want these models to be. So I think the only thing that we can be able to do is interpret the things we're seeing or the things we're paying attention to and then help us in using that information to do things. Yeah.Shawn Wang [00:19:05]: I think motion, you know, I still want to shout out, I think Gemini, still the only native video understanding model that's out there. So I use it for YouTube all the time. Nice.Jeff Dean [00:19:15]: Yeah. Yeah. I mean, it's actually, I think people kind of are not necessarily aware of what the Gemini models can actually do. Yeah. Like I have an example I've used in one of my talks. It had like, it was like a YouTube highlight video of 18 memorable sports moments across the last 20 years or something. So it has like Michael Jordan hitting some jump shot at the end of the finals and, you know, some soccer goals and things like that. And you can literally just give it the video and say, can you please make me a table of what all these different events are? What when the date is when they happened? And a short description. And so you get like now an 18 row table of that information extracted from the video, which is, you know, not something most people think of as like a turn video into sequel like table.Alessio Fanelli [00:20:11]: Has there been any discussion inside of Google of like, you mentioned tending to the whole internet, right? Google, it's almost built because a human cannot tend to the whole internet and you need some sort of ranking to find what you need. Yep. That ranking is like much different for an LLM because you can expect a person to look at maybe the first five, six links in a Google search versus for an LLM. Should you expect to have 20 links that are highly relevant? Like how do you internally figure out, you know, how do we build the AI mode that is like maybe like much broader search and span versus like the more human one? Yeah.Jeff Dean [00:20:47]: I mean, I think even pre-language model based work, you know, our ranking systems would be built to start. I mean, I think even pre-language model based work, you know, our ranking systems would be built to start. With a giant number of web pages in our index, many of them are not relevant. So you identify a subset of them that are relevant with very lightweight kinds of methods. You know, you're down to like 30,000 documents or something. And then you gradually refine that to apply more and more sophisticated algorithms and more and more sophisticated sort of signals of various kinds in order to get down to ultimately what you show, which is, you know, the final 10 results or, you know, 10 results plus. Other kinds of information. And I think an LLM based system is not going to be that dissimilar, right? You're going to attend to trillions of tokens, but you're going to want to identify, you know, what are the 30,000 ish documents that are with the, you know, maybe 30 million interesting tokens. And then how do you go from that into what are the 117 documents I really should be paying attention to in order to carry out the tasks that the user has asked? And I think, you know, you can imagine systems where you have, you know, a lot of highly parallel processing to identify those initial 30,000 candidates, maybe with very lightweight kinds of models. Then you have some system that sort of helps you narrow down from 30,000 to the 117 with maybe a little bit more sophisticated model or set of models. And then maybe the final model is the thing that looks. So the 117 things that might be your most capable model. So I think it has to, it's going to be some system like that, that is really enables you to give the illusion of attending to trillions of tokens. Sort of the way Google search gives you, you know, not the illusion, but you are searching the internet, but you're finding, you know, a very small subset of things that are, that are relevant.Shawn Wang [00:22:47]: Yeah. I often tell a lot of people that are not steeped in like Google search history that, well, you know, like Bert was. Like he was like basically immediately inside of Google search and that improves results a lot, right? Like I don't, I don't have any numbers off the top of my head, but like, I'm sure you guys, that's obviously the most important numbers to Google. Yeah.Jeff Dean [00:23:08]: I mean, I think going to an LLM based representation of text and words and so on enables you to get out of the explicit hard notion of, of particular words having to be on the page, but really getting at the notion of this topic of this page or this page. Paragraph is highly relevant to this query. Yeah.Shawn Wang [00:23:28]: I don't think people understand how much LLMs have taken over all these very high traffic system, very high traffic. Yeah. Like it's Google, it's YouTube. YouTube has this like semantics ID thing where it's just like every token or every item in the vocab is a YouTube video or something that predicts the video using a code book, which is absurd to me for YouTube size.Jeff Dean [00:23:50]: And then most recently GROK also for, for XAI, which is like, yeah. I mean, I'll call out even before LLMs were used extensively in search, we put a lot of emphasis on softening the notion of what the user actually entered into the query.Shawn Wang [00:24:06]: So do you have like a history of like, what's the progression? Oh yeah.Jeff Dean [00:24:09]: I mean, I actually gave a talk in, uh, I guess, uh, web search and data mining conference in 2009, uh, where we never actually published any papers about the origins of Google search, uh, sort of, but we went through sort of four or five or six. generations, four or five or six generations of, uh, redesigning of the search and retrieval system, uh, from about 1999 through 2004 or five. And that talk is really about that evolution. And one of the things that really happened in 2001 was we were sort of working to scale the system in multiple dimensions. So one is we wanted to make our index bigger, so we could retrieve from a larger index, which always helps your quality in general. Uh, because if you don't have the page in your index, you're going to not do well. Um, and then we also needed to scale our capacity because we were, our traffic was growing quite extensively. Um, and so we had, you know, a sharded system where you have more and more shards as the index grows, you have like 30 shards. And then if you want to double the index size, you make 60 shards so that you can bound the latency by which you respond for any particular user query. Um, and then as traffic grows, you add, you add more and more replicas of each of those. And so we eventually did the math that realized that in a data center where we had say 60 shards and, um, you know, 20 copies of each shard, we now had 1200 machines, uh, with disks. And we did the math and we're like, Hey, one copy of that index would actually fit in memory across 1200 machines. So in 2001, we introduced, uh, we put our entire index in memory and what that enabled from a quality perspective was amazing. Um, and so we had more and more replicas of each of those. Before you had to be really careful about, you know, how many different terms you looked at for a query, because every one of them would involve a disk seek on every one of the 60 shards. And so you, as you make your index bigger, that becomes even more inefficient. But once you have the whole index in memory, it's totally fine to have 50 terms you throw into the query from the user's original three or four word query, because now you can add synonyms like restaurant and restaurants and cafe and, uh, you know, things like that. Uh, bistro and all these things. And you can suddenly start, uh, sort of really, uh, getting at the meaning of the word as opposed to the exact semantic form the user typed in. And that was, you know, 2001, very much pre LLM, but really it was about softening the, the strict definition of what the user typed in order to get at the meaning.Alessio Fanelli [00:26:47]: What are like principles that you use to like design the systems, especially when you have, I mean, in 2001, the internet is like. Doubling, tripling every year in size is not like, uh, you know, and I think today you kind of see that with LLMs too, where like every year the jumps in size and like capabilities are just so big. Are there just any, you know, principles that you use to like, think about this? Yeah.Jeff Dean [00:27:08]: I mean, I think, uh, you know, first, whenever you're designing a system, you want to understand what are the sort of design parameters that are going to be most important in designing that, you know? So, you know, how many queries per second do you need to handle? How big is the internet? How big is the index you need to handle? How much data do you need to keep for every document in the index? How are you going to look at it when you retrieve things? Um, what happens if traffic were to double or triple, you know, will that system work well? And I think a good design principle is you're going to want to design a system so that the most important characteristics could scale by like factors of five or 10, but probably not beyond that because often what happens is if you design a system for X. And something suddenly becomes a hundred X, that would enable a very different point in the design space that would not make sense at X. But all of a sudden at a hundred X makes total sense. So like going from a disk space index to a in memory index makes a lot of sense once you have enough traffic, because now you have enough replicas of the sort of state on disk that those machines now actually can hold, uh, you know, a full copy of the, uh, index and memory. Yeah. And that all of a sudden enabled. A completely different design that wouldn't have been practical before. Yeah. Um, so I'm, I'm a big fan of thinking through designs in your head, just kind of playing with the design space a little before you actually do a lot of writing of code. But, you know, as you said, in the early days of Google, we were growing the index, uh, quite extensively. We were growing the update rate of the index. So the update rate actually is the parameter that changed the most. Surprising. So it used to be once a month.Shawn Wang [00:28:55]: Yeah.Jeff Dean [00:28:56]: And then we went to a system that could update any particular page in like sub one minute. Okay.Shawn Wang [00:29:02]: Yeah. Because this is a competitive advantage, right?Jeff Dean [00:29:04]: Because all of a sudden news related queries, you know, if you're, if you've got last month's news index, it's not actually that useful for.Shawn Wang [00:29:11]: News is a special beast. Was there any, like you could have split it onto a separate system.Jeff Dean [00:29:15]: Well, we did. We launched a Google news product, but you also want news related queries that people type into the main index to also be sort of updated.Shawn Wang [00:29:23]: So, yeah, it's interesting. And then you have to like classify whether the page is, you have to decide which pages should be updated and what frequency. Oh yeah.Jeff Dean [00:29:30]: There's a whole like, uh, system behind the scenes that's trying to decide update rates and importance of the pages. So even if the update rate seems low, you might still want to recrawl important pages quite often because, uh, the likelihood they change might be low, but the value of having updated is high.Shawn Wang [00:29:50]: Yeah, yeah, yeah, yeah. Uh, well, you know, yeah. This, uh, you know, mention of latency and, and saving things to this reminds me of one of your classics, which I have to bring up, which is latency numbers. Every programmer should know, uh, was there a, was it just a, just a general story behind that? Did you like just write it down?Jeff Dean [00:30:06]: I mean, this has like sort of eight or 10 different kinds of metrics that are like, how long does a cache mistake? How long does branch mispredict take? How long does a reference domain memory take? How long does it take to send, you know, a packet from the U S to the Netherlands or something? Um,Shawn Wang [00:30:21]: why Netherlands, by the way, or is it, is that because of Chrome?Jeff Dean [00:30:25]: Uh, we had a data center in the Netherlands, um, so, I mean, I think this gets to the point of being able to do the back of the envelope calculations. So these are sort of the raw ingredients of those, and you can use them to say, okay, well, if I need to design a system to do image search and thumb nailing or something of the result page, you know, how, what I do that I could pre-compute the image thumbnails. I could like. Try to thumbnail them on the fly from the larger images. What would that do? How much dis bandwidth than I need? How many des seeks would I do? Um, and you can sort of actually do thought experiments in, you know, 30 seconds or a minute with the sort of, uh, basic, uh, basic numbers at your fingertips. Uh, and then as you sort of build software using higher level libraries, you kind of want to develop the same intuitions for how long does it take to, you know, look up something in this particular kind of.Shawn Wang [00:31:21]: I'll see you next time.Shawn Wang [00:31:51]: Which is a simple byte conversion. That's nothing interesting. I wonder if you have any, if you were to update your...Jeff Dean [00:31:58]: I mean, I think it's really good to think about calculations you're doing in a model, either for training or inference.Jeff Dean [00:32:09]: Often a good way to view that is how much state will you need to bring in from memory, either like on-chip SRAM or HBM from the accelerator. Attached memory or DRAM or over the network. And then how expensive is that data motion relative to the cost of, say, an actual multiply in the matrix multiply unit? And that cost is actually really, really low, right? Because it's order, depending on your precision, I think it's like sub one picodule.Shawn Wang [00:32:50]: Oh, okay. You measure it by energy. Yeah. Yeah.Jeff Dean [00:32:52]: Yeah. I mean, it's all going to be about energy and how do you make the most energy efficient system. And then moving data from the SRAM on the other side of the chip, not even off the off chip, but on the other side of the same chip can be, you know, a thousand picodules. Oh, yeah. And so all of a sudden, this is why your accelerators require batching. Because if you move, like, say, the parameter of a model from SRAM on the, on the chip into the multiplier unit, that's going to cost you a thousand picodules. So you better make use of that, that thing that you moved many, many times with. So that's where the batch dimension comes in. Because all of a sudden, you know, if you have a batch of 256 or something, that's not so bad. But if you have a batch of one, that's really not good.Shawn Wang [00:33:40]: Yeah. Yeah. Right.Jeff Dean [00:33:41]: Because then you paid a thousand picodules in order to do your one picodule multiply.Shawn Wang [00:33:46]: I have never heard an energy-based analysis of batching.Jeff Dean [00:33:50]: Yeah. I mean, that's why people batch. Yeah. Ideally, you'd like to use batch size one because the latency would be great.Shawn Wang [00:33:56]: The best latency.Jeff Dean [00:33:56]: But the energy cost and the compute cost inefficiency that you get is quite large. So, yeah.Shawn Wang [00:34:04]: Is there a similar trick like, like, like you did with, you know, putting everything in memory? Like, you know, I think obviously NVIDIA has caused a lot of waves with betting very hard on SRAM with Grok. I wonder if, like, that's something that you already saw with, with the TPUs, right? Like that, that you had to. Uh, to serve at your scale, uh, you probably sort of saw that coming. Like what, what, what hardware, uh, innovations or insights were formed because of what you're seeing there?Jeff Dean [00:34:33]: Yeah. I mean, I think, you know, TPUs have this nice, uh, sort of regular structure of 2D or 3D meshes with a bunch of chips connected. Yeah. And each one of those has HBM attached. Um, I think for serving some kinds of models, uh, you know, you, you pay a lot higher cost. Uh, and time latency, um, bringing things in from HBM than you do bringing them in from, uh, SRAM on the chip. So if you have a small enough model, you can actually do model parallelism, spread it out over lots of chips and you actually get quite good throughput improvements and latency improvements from doing that. And so you're now sort of striping your smallish scale model over say 16 or 64 chips. Uh, but as if you do that and it all fits in. In SRAM, uh, that can be a big win. So yeah, that's not a surprise, but it is a good technique.Alessio Fanelli [00:35:27]: Yeah. What about the TPU design? Like how much do you decide where the improvements have to go? So like, this is like a good example of like, is there a way to bring the thousand picojoules down to 50? Like, is it worth designing a new chip to do that? The extreme is like when people say, oh, you should burn the model on the ASIC and that's kind of like the most extreme thing. How much of it? Is it worth doing an hardware when things change so quickly? Like what was the internal discussion? Yeah.Jeff Dean [00:35:57]: I mean, we, we have a lot of interaction between say the TPU chip design architecture team and the sort of higher level modeling, uh, experts, because you really want to take advantage of being able to co-design what should future TPUs look like based on where we think the sort of ML research puck is going, uh, in some sense, because, uh, you know, as a hardware designer for ML and in particular, you're trying to design a chip starting today and that design might take two years before it even lands in a data center. And then it has to sort of be a reasonable lifetime of the chip to take you three, four or five years. So you're trying to predict two to six years out where, what ML computations will people want to run two to six years out in a very fast changing field. And so having people with interest. Interesting ML research ideas of things we think will start to work in that timeframe or will be more important in that timeframe, uh, really enables us to then get, you know, interesting hardware features put into, you know, TPU N plus two, where TPU N is what we have today.Shawn Wang [00:37:10]: Oh, the cycle time is plus two.Jeff Dean [00:37:12]: Roughly. Wow. Because, uh, I mean, sometimes you can squeeze some changes into N plus one, but, you know, bigger changes are going to require the chip. Yeah. Design be earlier in its lifetime design process. Um, so whenever we can do that, it's generally good. And sometimes you can put in speculative features that maybe won't cost you much chip area, but if it works out, it would make something, you know, 10 times as fast. And if it doesn't work out, well, you burned a little bit of tiny amount of your chip area on that thing, but it's not that big a deal. Uh, sometimes it's a very big change and we want to be pretty sure this is going to work out. So we'll do like lots of carefulness. Uh, ML experimentation to show us, uh, this is actually the, the way we want to go. Yeah.Alessio Fanelli [00:37:58]: Is there a reverse of like, we already committed to this chip design so we can not take the model architecture that way because it doesn't quite fit?Jeff Dean [00:38:06]: Yeah. I mean, you, you definitely have things where you're going to adapt what the model architecture looks like so that they're efficient on the chips that you're going to have for both training and inference of that, of that, uh, generation of model. So I think it kind of goes both ways. Um, you know, sometimes you can take advantage of, you know, lower precision things that are coming in a future generation. So you can, might train it at that lower precision, even if the current generation doesn't quite do that. Mm.Shawn Wang [00:38:40]: Yeah. How low can we go in precision?Jeff Dean [00:38:43]: Because people are saying like ternary is like, uh, yeah, I mean, I'm a big fan of very low precision because I think that gets, that saves you a tremendous amount of time. Right. Because it's picojoules per bit that you're transferring and reducing the number of bits is a really good way to, to reduce that. Um, you know, I think people have gotten a lot of luck, uh, mileage out of having very low bit precision things, but then having scaling factors that apply to a whole bunch of, uh, those, those weights. Scaling. How does it, how does it, okay.Shawn Wang [00:39:15]: Interesting. You, so low, low precision, but scaled up weights. Yeah. Huh. Yeah. Never considered that. Yeah. Interesting. Uh, w w while we're on this topic, you know, I think there's a lot of, um, uh, this, the concept of precision at all is weird when we're sampling, you know, uh, we just, at the end of this, we're going to have all these like chips that I'll do like very good math. And then we're just going to throw a random number generator at the start. So, I mean, there's a movement towards, uh, energy based, uh, models and processors. I'm just curious if you've, obviously you've thought about it, but like, what's your commentary?Jeff Dean [00:39:50]: Yeah. I mean, I think. There's a bunch of interesting trends though. Energy based models is one, you know, diffusion based models, which don't sort of sequentially decode tokens is another, um, you know, speculative decoding is a way that you can get sort of an equivalent, very small.Shawn Wang [00:40:06]: Draft.Jeff Dean [00:40:07]: Batch factor, uh, for like you predict eight tokens out and that enables you to sort of increase the effective batch size of what you're doing by a factor of eight, even, and then you maybe accept five or six of those tokens. So you get. A five, a five X improvement in the amortization of moving weights, uh, into the multipliers to do the prediction for the, the tokens. So these are all really good techniques and I think it's really good to look at them from the lens of, uh, energy, real energy, not energy based models, um, and, and also latency and throughput, right? If you look at things from that lens, that sort of guides you to. Two solutions that are gonna be, uh, you know, better from, uh, you know, being able to serve larger models or, you know, equivalent size models more cheaply and with lower latency.Shawn Wang [00:41:03]: Yeah. Well, I think, I think I, um, it's appealing intellectually, uh, haven't seen it like really hit the mainstream, but, um, I do think that, uh, there's some poetry in the sense that, uh, you know, we don't have to do, uh, a lot of shenanigans if like we fundamentally. Design it into the hardware. Yeah, yeah.Jeff Dean [00:41:23]: I mean, I think there's still a, there's also sort of the more exotic things like analog based, uh, uh, computing substrates as opposed to digital ones. Uh, I'm, you know, I think those are super interesting cause they can be potentially low power. Uh, but I think you often end up wanting to interface that with digital systems and you end up losing a lot of the power advantages in the digital to analog and analog to digital conversions. You end up doing, uh, at the sort of boundaries. And periphery of that system. Um, I still think there's a tremendous distance we can go from where we are today in terms of energy efficiency with sort of, uh, much better and specialized hardware for the models we care about.Shawn Wang [00:42:05]: Yeah.Alessio Fanelli [00:42:06]: Um, any other interesting research ideas that you've seen, or like maybe things that you cannot pursue a Google that you would be interested in seeing researchers take a step at, I guess you have a lot of researchers. Yeah, I guess you have enough, but our, our research.Jeff Dean [00:42:21]: Our research portfolio is pretty broad. I would say, um, I mean, I think, uh, in terms of research directions, there's a whole bunch of, uh, you know, open problems and how do you make these models reliable and able to do much longer, kind of, uh, more complex tasks that have lots of subtasks. How do you orchestrate, you know, maybe one model that's using other models as tools in order to sort of build, uh, things that can accomplish, uh, you know, much more. Yeah. Significant pieces of work, uh, collectively, then you would ask a single model to do. Um, so that's super interesting. How do you get more verifiable, uh, you know, how do you get RL to work for non-verifiable domains? I think it's a pretty interesting open problem because I think that would broaden out the capabilities of the models, the improvements that you're seeing in both math and coding. Uh, if we could apply those to other less verifiable domains, because we've come up with RL techniques that actually enable us to do that. Uh, effectively, that would, that would really make the models improve quite a lot. I think.Alessio Fanelli [00:43:26]: I'm curious, like when we had Noam Brown on the podcast, he said, um, they already proved you can do it with deep research. Um, you kind of have it with AI mode in a way it's not verifiable. I'm curious if there's any thread that you think is interesting there. Like what is it? Both are like information retrieval of JSON. So I wonder if it's like the retrieval is like the verifiable part. That you can score or what are like, yeah, yeah. How, how would you model that, that problem?Jeff Dean [00:43:55]: Yeah. I mean, I think there are ways of having other models that can evaluate the results of what a first model did, maybe even retrieving. Can you have another model that says, is this things, are these things you retrieved relevant? Or can you rate these 2000 things you retrieved to assess which ones are the 50 most relevant or something? Um, I think those kinds of techniques are actually quite effective. Sometimes I can even be the same model, just prompted differently to be a, you know, a critic as opposed to a, uh, actual retrieval system. Yeah.Shawn Wang [00:44:28]: Um, I do think like there, there is that, that weird cliff where like, it feels like we've done the easy stuff and then now it's, but it always feels like that every year. It's like, oh, like we know, we know, and the next part is super hard and nobody's figured it out. And, uh, exactly with this RLVR thing where like everyone's talking about, well, okay, how do we. the next stage of the non-verifiable stuff. And everyone's like, I don't know, you know, Ellen judge.Jeff Dean [00:44:56]: I mean, I feel like the nice thing about this field is there's lots and lots of smart people thinking about creative solutions to some of the problems that we all see. Uh, because I think everyone sort of sees that the models, you know, are great at some things and they fall down around the edges of those things and, and are not as capable as we'd like in those areas. And then coming up with good techniques and trying those. And seeing which ones actually make a difference is sort of what the whole research aspect of this field is, is pushing forward. And I think that's why it's super interesting. You know, if you think about two years ago, we were struggling with GSM, eight K problems, right? Like, you know, Fred has two rabbits. He gets three more rabbits. How many rabbits does he have? That's a pretty far cry from the kinds of mathematics that the models can, and now you're doing IMO and Erdos problems in pure language. Yeah. Yeah. Pure language. So that is a really, really amazing jump in capabilities in, you know, in a year and a half or something. And I think, um, for other areas, it'd be great if we could make that kind of leap. Uh, and you know, we don't exactly see how to do it for some, some areas, but we do see it for some other areas and we're going to work hard on making that better. Yeah.Shawn Wang [00:46:13]: Yeah.Alessio Fanelli [00:46:14]: Like YouTube thumbnail generation. That would be very helpful. We need that. That would be AGI. We need that.Shawn Wang [00:46:20]: That would be. As far as content creators go.Jeff Dean [00:46:22]: I guess I'm not a YouTube creator, so I don't care that much about that problem, but I guess, uh, many people do.Shawn Wang [00:46:27]: It does. Yeah. It doesn't, it doesn't matter. People do judge books by their covers as it turns out. Um, uh, just to draw a bit on the IMO goal. Um, I'm still not over the fact that a year ago we had alpha proof and alpha geometry and all those things. And then this year we were like, screw that we'll just chuck it into Gemini. Yeah. What's your reflection? Like, I think this, this question about. Like the merger of like symbolic systems and like, and, and LMS, uh, was a very much core belief. And then somewhere along the line, people would just said, Nope, we'll just all do it in the LLM.Jeff Dean [00:47:02]: Yeah. I mean, I think it makes a lot of sense to me because, you know, humans manipulate symbols, but we probably don't have like a symbolic representation in our heads. Right. We have some distributed representation that is neural net, like in some way of lots of different neurons. And activation patterns firing when we see certain things and that enables us to reason and plan and, you know, do chains of thought and, you know, roll them back now that, that approach for solving the problem doesn't seem like it's going to work. I'm going to try this one. And, you know, in a lot of ways we're emulating what we intuitively think, uh, is happening inside real brains in neural net based models. So it never made sense to me to have like completely separate. Uh, discrete, uh, symbolic things, and then a completely different way of, of, uh, you know, thinking about those things.Shawn Wang [00:47:59]: Interesting. Yeah. Uh, I mean, it's maybe seems obvious to you, but it wasn't obvious to me a year ago. Yeah.Jeff Dean [00:48:06]: I mean, I do think like that IMO with, you know, translating to lean and using lean and then the next year and also a specialized geometry model. And then this year switching to a single unified model. That is roughly the production model with a little bit more inference budget, uh, is actually, you know, quite good because it shows you that the capabilities of that general model have improved dramatically and, and now you don't need the specialized model. This is actually sort of very similar to the 2013 to 16 era of machine learning, right? Like it used to be, people would train separate models for lots of different, each different problem, right? I have, I want to recognize street signs and something. So I train a street sign. Recognition recognition model, or I want to, you know, decode speech recognition. I have a speech model, right? I think now the era of unified models that do everything is really upon us. And the question is how well do those models generalize to new things they've never been asked to do and they're getting better and better.Shawn Wang [00:49:10]: And you don't need domain experts. Like one of my, uh, so I interviewed ETA who was on, who was on that team. Uh, and he was like, yeah, I, I don't know how they work. I don't know where the IMO competition was held. I don't know the rules of it. I just trained the models, the training models. Yeah. Yeah. And it's kind of interesting that like people with these, this like universal skill set of just like machine learning, you just give them data and give them enough compute and they can kind of tackle any task, which is the bitter lesson, I guess. I don't know. Yeah.Jeff Dean [00:49:39]: I mean, I think, uh, general models, uh, will win out over specialized ones in most cases.Shawn Wang [00:49:45]: Uh, so I want to push there a bit. I think there's one hole here, which is like, uh. There's this concept of like, uh, maybe capacity of a model, like abstractly a model can only contain the number of bits that it has. And, uh, and so it, you know, God knows like Gemini pro is like one to 10 trillion parameters. We don't know, but, uh, the Gemma models, for example, right? Like a lot of people want like the open source local models that are like that, that, that, and, and, uh, they have some knowledge, which is not necessary, right? Like they can't know everything like, like you have the. The luxury of you have the big model and big model should be able to capable of everything. But like when, when you're distilling and you're going down to the small models, you know, you're actually memorizing things that are not useful. Yeah. And so like, how do we, I guess, do we want to extract that? Can we, can we divorce knowledge from reasoning, you know?Jeff Dean [00:50:38]: Yeah. I mean, I think you do want the model to be most effective at reasoning if it can retrieve things, right? Because having the model devote precious parameter space. To remembering obscure facts that could be looked up is actually not the best use of that parameter space, right? Like you might prefer something that is more generally useful in more settings than this obscure fact that it has. Um, so I think that's always attention at the same time. You also don't want your model to be kind of completely detached from, you know, knowing stuff about the world, right? Like it's probably useful to know how long the golden gate be. Bridges just as a general sense of like how long are bridges, right? And, uh, it should have that kind of knowledge. It maybe doesn't need to know how long some teeny little bridge in some other more obscure part of the world is, but, uh, it does help it to have a fair bit of world knowledge and the bigger your model is, the more you can have. Uh, but I do think combining retrieval with sort of reasoning and making the model really good at doing multiple stages of retrieval. Yeah.Shawn Wang [00:51:49]: And reasoning through the intermediate retrieval results is going to be a, a pretty effective way of making the model seem much more capable, because if you think about, say, a personal Gemini, yeah, right?Jeff Dean [00:52:01]: Like we're not going to train Gemini on my email. Probably we'd rather have a single model that, uh, we can then use and use being able to retrieve from my email as a tool and have the model reason about it and retrieve from my photos or whatever, uh, and then make use of that and have multiple. Um, you know, uh, stages of interaction. that makes sense.Alessio Fanelli [00:52:24]: Do you think the vertical models are like, uh, interesting pursuit? Like when people are like, oh, we're building the best healthcare LLM, we're building the best law LLM, are those kind of like short-term stopgaps or?Jeff Dean [00:52:37]: No, I mean, I think, I think vertical models are interesting. Like you want them to start from a pretty good base model, but then you can sort of, uh, sort of viewing them, view them as enriching the data. Data distribution for that particular vertical domain for healthcare, say, um, we're probably not going to train or for say robotics. We're probably not going to train Gemini on all possible robotics data. We, you could train it on because we want it to have a balanced set of capabilities. Um, so we'll expose it to some robotics data, but if you're trying to build a really, really good robotics model, you're going to want to start with that and then train it on more robotics data. And then maybe that would. It's multilingual translation capability, but improve its robotics capabilities. And we're always making these kind of, uh, you know, trade-offs in the data mix that we train the base Gemini models on. You know, we'd love to include data from 200 more languages and as much data as we have for those languages, but that's going to displace some other capabilities of the model. It won't be as good at, um, you know, Pearl programming, you know, it'll still be good at Python programming. Cause we'll include it. Enough. Of that, but there's other long tail computer languages or coding capabilities that it may suffer on or multi, uh, multimodal reasoning capabilities may suffer. Cause we didn't get to expose it to as much data there, but it's really good at multilingual things. So I, I think some combination of specialized models, maybe more modular models. So it'd be nice to have the capability to have those 200 languages, plus this awesome robotics model, plus this awesome healthcare, uh, module that all can be knitted together to work in concert and called upon in different circumstances. Right? Like if I have a health related thing, then it should enable using this health module in conjunction with the main base model to be even better at those kinds of things. Yeah.Shawn Wang [00:54:36]: Installable knowledge. Yeah.Jeff Dean [00:54:37]: Right.Shawn Wang [00:54:38]: Just download as a, as a package.Jeff Dean [00:54:39]: And some of that installable stuff can come from retrieval, but some of it probably should come from preloaded training on, you know, uh, a hundred billion tokens or a trillion tokens of health data. Yeah.Shawn Wang [00:54:51]: And for listeners, I think, uh, I will highlight the Gemma three end paper where they, there was a little bit of that, I think. Yeah.Alessio Fanelli [00:54:56]: Yeah. I guess the question is like, how many billions of tokens do you need to outpace the frontier model improvements? You know, it's like, if I have to make this model better healthcare and the main. Gemini model is still improving. Do I need 50 billion tokens? Can I do it with a hundred, if I need a trillion healthcare tokens, it's like, they're probably not out there that you don't have, you know, I think that's really like the.Jeff Dean [00:55:21]: Well, I mean, I think healthcare is a particularly challenging domain, so there's a lot of healthcare data that, you know, we don't have access to appropriately, but there's a lot of, you know, uh, healthcare organizations that want to train models on their own data. That is not public healthcare data, uh, not public health. But public healthcare data. Um, so I think there are opportunities there to say, partner with a large healthcare organization and train models for their use that are going to be, you know, more bespoke, but probably, uh, might be better than a general model trained on say, public data. Yeah.Shawn Wang [00:55:58]: Yeah. I, I believe, uh, by the way, also this is like somewhat related to the language conversation. Uh, I think one of your, your favorite examples was you can put a low resource language in the context and it just learns. Yeah.Jeff Dean [00:56:09]: Oh, yeah, I think the example we used was Calamon, which is truly low resource because it's only spoken by, I think 120 people in the world and there's no written text.Shawn Wang [00:56:20]: So, yeah. So you can just do it that way. Just put it in the context. Yeah. Yeah. But I think your whole data set in the context, right.Jeff Dean [00:56:27]: If you, if you take a language like, uh, you know, Somali or something, there is a fair bit of Somali text in the world that, uh, or Ethiopian Amharic or something, um, you know, we probably. Yeah. Are not putting all the data from those languages into the Gemini based training. We put some of it, but if you put more of it, you'll improve the capabilities of those models.Shawn Wang [00:56:49]: Yeah.Jeff Dean [00:56:49]:

Wen Quan Cheong, co-manager of Mawer's emerging markets equity strategy, outlines four major themes shaping the opportunity set today. First, the "picks and shovels" of AI: upstream enablers such as advanced chip manufacturers, memory makers, and specialized chip-testing firms that are benefiting from structural bottlenecks in the AI supply chain. Second, companies that are actually converting AI investment into higher returns on capital. Third, the "Great Supply Chain Reshuffle," where national security concerns, tariffs, and "China plus one" strategies are driving a reconfiguration of strategic manufacturing infrastructure across Asia and the U.S. And finally, a broader universe of less obvious EM stories that illustrate how opportunity is evolving across regions and sectors as these forces play out.   Highlights: Why upstream AI enablers are seeing such powerful earnings leverage: how capacity cuts, equipment bottlenecks, and surging demand for DRAM, HBM, and NAND have flipped the memory market from oversupplied to structurally tight. What it takes for companies to truly convert AI investment into sustainable returns on invested capital, and why early, well-run adopters may enjoy a multi year edge. How shifting geopolitics, U.S. tariffs, and national security concerns are driving a "Great Supply Chain Reshuffle," from TSMC-linked clean room specialists like Actor Group supporting new fabs to Chinese manufacturers using their domestic scale and integration to expand overseas. Why emerging markets are more than just China and tech, with examples ranging from Saudi insurance aggregation and Vietnamese pharmacies to ship maintenance businesses with recurring revenues.   Host: Rob Campbell, CFA Institutional Portfolio Manager Guest: Wen Quan Cheong, CFA Portfolio Manager   This episode is available for download anywhere you get your podcasts. Founded in 1974, Mawer Investment Management Ltd. (pronounced "more") is a privately owned independent investment firm managing assets for institutional and individual investors. Mawer employs over 250 people in Canada, U.S., and Singapore. Visit Mawer at https://www.mawer.com. Follow us on social: LinkedIn - https://www.linkedin.com/company/mawer-investment-management/ Instagram - https://www.instagram.com/mawerinvestmentmanagement/

For the 100th episode of Astonishing Healthcare, we welcomed AJ Loiacono, our co-founder and CEO, back to the show for a lively discussion about the evolution of our industry and business. What started as a transparent pharmacy benefits manager (PBM) in the "age of indifference" is now a more comprehensive health benefits manager (HBM), and we've entered the "era of acceptance." It's been an incredible 8+ years of growth, fueled by innovation and an unwavering commitment to our clients and delivering on our mission: to build the infrastructure our country needs to deliver the healthcare we deserve. But we had to endure an "age of confusion" to get here!AJ explains why traditional healthcare giants are facing a "BlackBerry moment" - trying to emulate a conflict-free challenger when "it's already too late." The balance of power is shifting away from the traditional PBMs, as the industry now demands full transparency - buyers of health benefits today are smarter than ever before. We also discuss how and why the U.S. wastes [at least] a trillion dollars annually by trying to deliver care using inefficient, fragmented systems; we built the infrastructure to stop it. This episode isn't just a retrospective; it's a blueprint of sorts, and we've got the cultural DNA required to bring about sustainable change (vs. just daydreaming about it). Related ContentReplay - Unifying Medical and Pharmacy Benefits: The Blueprint for Better Employee Health and WellnessJudi Health's Capital Rx Surpasses Five Million Contracted PBM Lives as America's Largest Employers, Unions, and Leading Health Systems Evolve Their Health Benefits StrategiesAH095 - What's in Store for the New Year? A Special Round-Robin Episode of Astonishing HealthcareHealth Benefits 101: Service Excellence & Scaling an Award-Winning Call Center ModelFor more information about Judi Health and this episode, please visit Judi Health - Insights.

In this episode Katherine Forrest and Scott Caravello take us down “memory lane” to explain the importance of high bandwidth memory (HBM) and RAM to AI development. Our hosts also give us a rundown of potential challenges ahead, unpacking developments in the market for memory, including plans for additional capacity and lobster-style RAM pricing. ## Learn More About Paul, Weiss's Artificial Intelligence practice: https://www.paulweiss.com/industries/artificial-intelligence

See my $350,000+ Stock Portfolio: https://www.patreon.com/citizenoftheyear/postsJoin the discord: https://discord.gg/Gq8hGbg2CqCheck out these AMAZING Deals: https://amzn.to/3NGmBPTMicron stock has surged because the company has become a key supplier of memory chips for AI, especially high-bandwidth memory (HBM), which is already sold out through 2026. Strong AI demand, record earnings, and unusually high profit margins have caused investors to view Micron less as a cyclical memory stock and more as critical AI infrastructure. The big risk for Micron stock long term is whether competitors like Samsung flood the market and drive prices down, or if memory truly stays essential to the AI economy.Check out my favorite research tool Seeking Alpha! Premium: https://link.seekingalpha.com/3B2L85W/4G6SHH/Alpha Picks: https://www.sahg6dtr.com/3B2L85W/J8P3N/Disclaimer:This is not financial advice and I am not a licensed financial advisor. Always do your own research before investing and work with a licensed financial advisor. These are my opinions for informational purposes only and not to be taken as investing advice. Some of the links on this page are affiliate links, meaning, at no additional cost to you, I may earn a commission if you click through and make a purchase and/or subscribe. As an Amazon Associate, I earn from qualifying purchases. Affiliate commissions help fund videos like this one

Linktree: ⁠https://linktr.ee/Analytic⁠Join The Normandy For Additional Bonus Audio And Visual Content For All Things Nme+! Join Here: ⁠https://ow.ly/msoH50WCu0K⁠In the Notorious Mass Effect segment, Analytic Dreamz dives deep into the RAM Price Crisis (2025–2026), unpacking the key data, market drivers, and real consumer impact behind the dramatic surge in memory costs.RAM prices have skyrocketed into a sustained inflation cycle heading into 2026, fueled by explosive AI data center demand that prioritizes high-bandwidth memory (HBM) and diverts supply from consumer DRAM. Manufacturing bottlenecks, limited cleanroom capacity, and lithography constraints exacerbate the shortage, while major players like Micron exit consumer RAM sales (Crucial brand in December 2025) to focus on higher-margin AI segments. Samsung and SK hynix report massive profit surges amid the boom.DDR5 RAM has seen prices more than quadruple (+340–344%) since July 2025, with a +27% month-on-month jump from December to January 2026. DDR4 and older standards are rising even faster recently (+46% MoM in January), narrowing the gap with newer tech. ComputerBase's fixed-basket analysis confirms average prices have quadrupled versus September 2025, with Germany's retail tracking—Europe's largest PC hardware market—mirroring global trends, including growing secondary-market distortions.Secondary effects hit related components hard: SSDs up +79%, hard drives +53%, GPUs +14% (with street prices far exceeding MSRP on models like RTX 5070 Ti). Specific examples include 2TB NVMe drives jumping 60–159% and NAS HDDs doubling.Analyst forecasts from TrendForce and Omdia point to +50–60% DRAM contract price hikes in Q1 2026, following 40–70% YoY increases in 2025. PC shipments grew +9.2% in 2025 but face potential declines in 2026, while smartphone output forecasts drop ~20% for some brands, risking +30% price hikes or spec downgrades. Gaming consoles may see delays or higher launch prices.Apple's upgrade costs (e.g., $400 for 16GB→32GB) already outpace comparable DDR5 sticks, with M6 Macs potentially facing steeper hikes or supply delays if AI firms continue outbidding.The core takeaway: This AI-driven structural shift has quadrupled RAM prices in under six months, with volatility persisting through 2026. A plateau is the most optimistic scenario—no full reversal in sight. Analytic Dreamz breaks down the data, root causes, and widespread ripple effects across PCs, smartphones, and beyond.Support this podcast at — https://redcircle.com/analytic-dreamz-notorious-mass-effect/donationsPrivacy & Opt-Out: https://redcircle.com/privacy

NIO ships to 460K vehicles, Li Auto goes all-in on robots, memory crisis hits everyone. This is execution vs vision vs reality.NIO NWM UPDATE (460K+ VEHICLES):Major "human-like" driving update using closed-loop reinforcement learningLearns from REAL human driving, not just expertsBattery swap navigation: industry-first piloted driving to 2,000+ stationsShenji in-house chips (no NVIDIA delays)EXECUTION AT SCALE despite sales strugglesLI AUTO ROBOT PIVOT (LEAKED INTERNAL MEETING):Jan 26 all-hands: Li Xiang announces humanoid robot pushKey Points:2026 = last year to become top AI companyOnly 3 global companies will master foundation models + chips + OS + embodied intelligenceLi Auto will be oneRestructuring: cars + robots = "hardware ontology team"Aggressive hiring: "bring back employees who left for robot startups"Multiple robot R&D roles postedContext: Sales 500K (2024) → 400K (2025), -20%. Pure EV struggling. Is this genius or desperation?MEMORY CHIP CRISIS (AFFECTS ALL):DDR4/DDR5 prices +40-70%, adding 1,000-2,000 yuan per vehicleStats:Li Auto:

S&P futures is up +0.2% and pointing to a higher open today. Asian equities closed broadly higher Tuesday. SK Hynix has emerged as the exclusive supplier of HBM chips for Microsoft's Maia 200 AI chip, driving outsized gains in South Korea's markets. Japan's Nikkei was also higher on strength in exporters, while the Hang Seng led Greater China market gains. European markets are also higher in early trading. Companies Mentioned: Meta, SK Hynix, Ford, General Motors

Rambus (RMBS) has historically been known as an IP and patent powerhouse—but in 2026, the story has changed. With a massive memory chip shortage driving demand, Rambus is pivoting hard into becoming a fabless chip designer of memory interfaces. In this video, Nick and Kasey break down exactly how Rambus fits into the electronics manufacturing supply chain today. We analyze their transition from pure licensing to selling their own silicon (like memory interface chips for DDR5 and HBM), review their latest Q3 2025 financials, and discuss whether the current valuation makes sense for your portfolio.Join us on Discord with Semiconductor Insider, sign up on our website: www.chipstockinvestor.com/membershipSupercharge your analysis with AI! Get 15% of your membership with our special link here: https://fiscal.ai/csi/Sign Up For Our Newsletter: https://mailchi.mp/b1228c12f284/sign-up-landing-page-short-form

Memory shortages are all the rage in 2026. How should you play the AI data center supply crunch?We discussed this back in 2025, and now it is here: Memory shortages are hitting the AI data center supply chain across the board. But is this an AI bubble, or just a normal cyclical growth cycle? In this video, we break down the entire memory hierarchy—from ultra-fast on-chip SRAM to HBM and long-term storage—and give you the basket of companies to watch for each layer.We also discuss why Pure Storage is our top bet for secondary storage and how equipment suppliers like Lam Research could benefit as manufacturers race to expand capacity.Join us on Discord with Semiconductor Insider, sign up on our website: www.chipstockinvestor.com/membershipSupercharge your analysis with AI! Get 15% of your membership with our special link here: https://fiscal.ai/csi/Sign Up For Our Newsletter: https://mailchi.mp/b1228c12f284/sign-up-landing-page-short-formChapters:00:00 – Memory Shortages: Bubble vs. Cyclical Growth 02:13 – The AI Memory Hierarchy Explained (SRAM, DRAM, NAND) 04:59 – SRAM Stocks: Nvidia, AMD, & Synopsys 06:50 – Embedded Memory: Weebit Nano & MRAM players 07:46 – DRAM & HBM Leaders: SK Hynix, Micron, Samsung 09:00 – The NAND & HDD Resurgence (Seagate & WD) 11:00 – Why Pure Storage is a Top Bet 14:00 – The Fab Five & Lam Research OpportunityIf you found this video useful, please make sure to like and subscribe!*********************************************************Affiliate links that are sprinkled in throughout this video. If something catches your eye and you decide to buy it, we might earn a little coffee money. Thanks for helping us (Kasey) fuel our caffeine addiction!Content in this video is for general information or entertainment only and is not specific or individual investment advice. Forecasts and information presented may not develop as predicted and there is no guarantee any strategies presented will be successful. All investing involves risk, and you could lose some or all of your principal. #semiconductors #chips #investing #stocks #finance #financeeducation #silicon #artificialintelligence #ai #financeeducation #chipstocks #finance #stocks #investing #investor #financeeducation #stockmarket #chipstockinvestor #fablesschipdesign #chipmanufacturing #semiconductormanufacturing #semiconductorstocks Nick and Kasey own shares of Nvidia, Micron, Pure Storage, Sk hynix, Kioxia, Lam Research

En 2026, les appareils électroniques — smartphones, ordinateurs, tablettes, consoles ou objets connectés — vont coûter plus cher. L'une des raisons majeures, encore peu visible pour le grand public, est l'augmentation rapide du prix de la mémoire vive, la RAM. Et cette hausse est directement liée à l'explosion de l'intelligence artificielle.La RAM est un composant essentiel de tout appareil électronique. Elle permet de stocker temporairement les données utilisées par le processeur et conditionne la rapidité et la fluidité d'un système. Sans RAM, pas de multitâche, pas d'applications modernes, pas d'IA embarquée. Or, depuis deux ans, la demande mondiale de mémoire a changé de nature.Traditionnellement, la RAM était majoritairement destinée aux PC, aux smartphones et aux serveurs classiques. Désormais, les grandes entreprises de l'IA — OpenAI, Google, Microsoft, Meta, Amazon — consomment des quantités colossales de mémoire pour entraîner et faire fonctionner leurs modèles. Les serveurs d'IA utilisent des mémoires spécifiques, comme la HBM (High Bandwidth Memory), indispensables pour alimenter les puces de calcul de type GPU. Un seul serveur dédié à l'IA peut embarquer plusieurs centaines de gigaoctets de RAM, soit l'équivalent de dizaines, voire de centaines de smartphones.Selon plusieurs cabinets d'analyse, la demande en mémoire liée à l'IA progresse de plus de 40 % par an. En face, l'offre ne suit pas. Les fabricants de mémoire — Samsung, SK Hynix et Micron — ont volontairement limité leurs investissements après la crise de surproduction de 2022-2023. Résultat : en 2026, la production mondiale de DRAM devrait augmenter d'environ 15 à 16 %, bien moins que la demande.Ce déséquilibre a déjà un impact sur les prix. En 2025, les prix de la DRAM ont augmenté de plus de 50 %. Pour 2026, plusieurs prévisions évoquent une nouvelle hausse comprise entre 30 et 50 %, selon les segments. La mémoire HBM, très utilisée par l'IA, est encore plus sous tension, car elle mobilise davantage de silicium et des chaînes de production complexes, au détriment de la RAM “classique”.Or la RAM représente entre 10 et 20 % du coût de fabrication d'un PC ou d'un smartphone milieu et haut de gamme. Quand ce composant augmente fortement, les fabricants n'ont que deux options : réduire les performances ou augmenter les prix. De plus en plus, ils choisissent la seconde solution. Des hausses de prix sont déjà anticipées sur les PC et les smartphones dès 2026, avec une augmentation moyenne estimée entre 6 et 8 %.En résumé, l'essor fulgurant de l'intelligence artificielle accapare la mémoire mondiale. Et cette bataille invisible pour la RAM se traduira très concrètement, en 2026, par des appareils électroniques plus chers pour les consommateurs. Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.

US equity futures point to a mixed open, with Asian markets mostly lower and European equities trading slightly higher. Today focus is on continued risk aversion in US equities. Moreover, the global rate backdrop remains a headwind as markets digest a hawkish tilt in central bank expectations, with investors increasingly focused on upcoming US inflation data and jobless claims for confirmation on whether policy easing can resume next year. In addition, corporate developments remained in focus as Micron guided above expectations and lifted medium-term capital expenditure plans tied to HBM demand, offering selective support to memory-related names but failing to offset broader concerns around AI monetization, positioning fatigue, and elevated valuations.Companies Mentioned: OpenAI, Warner Bros. Discovery, lululemon athletica

Fifty years of Semicon Europa set a fitting backdrop for a conversation that feels both celebratory and unsentimental about the state of advanced packaging in Europe. We walk the floor in Munich and pull together a story that spans chemical metrology, panel plating, glass substrates, thermal materials, logistics resilience, and the push from R&D to production—plus a heartfelt goodbye.Dena Mitchell, Nova opens the curtain on chemical metrology for electroplating, showing how bath health drives TSV fill, hybrid bond grain structure, and environmental wins through longer bath life. Sally Ann Henry, ACM Research, explains why horizontal panel electroplating can deliver better uniformity than vertical as panel-level packaging grows. Thomas Uhrmann, EV Group zooms out to the strategy: Europe's strength in pilot lines and research consortia, the urgency to materialize large-scale packaging fabs, and how the EU Chips Act is knitting packaging into every node from photonics to logic.Henkel's Ram Trichur takes on thermals, from kilowatt-class data center processors with backside power delivery to mobile's shift from package-on-package to side-by-side for exposed die cooling, and the heat challenges inside HBM stacks. Comet's Isabella Drolz steps into glass panel territory with TGV inspection at 610 x 610 mm, aligning tools, standards, and timelines toward late-decade ramps. Martin Wynaendts van Resandt explains howLab14 brings agility with direct-write lithography for large substrates and optical interconnect masters—speeding iteration and trimming mask overhead as co-packaged optics advances. Jim Garstka, Shellback Semiconductor, talks about its Hydrozone product that is finding traction in photo mask cleaning.  We also get practical about moving all this innovation: Barry O'Dowd and Robin Knopf, of Kuehne+Nagel, detail how Europe's packaging supply chains remain global, and how sea-air blends can cut cost and time for non-sensitive, high-volume flows while building resilience against disruptions. ASE's Patricia MacLeod, Christophe Zinck, and Bradford Factor tie it together with automotive realities—centralized compute, heterogeneous integration, reliability constraints—and the enduring role of MEMS and sensors to feed the brain of the car.It's a grounded, forward-looking journey through the technologies and decisions that will determine whether Europe turns its R&D leadership into production momentum. Listen for clear takeaways, candid perspectives, and a final toast to the community that made the 3D InCites Podcast possible.If this conversation resonates, follow the show, share it with a colleague, and leave a review to help more listeners find it.Support the show