POPULARITY
9 episodes with Epyc
4 episodes with Epyc
11 episodes with Epyc
6 episodes with Epyc
4 episodes with Epyc
4 episodes with Epyc
8 episodes with Epyc
5 episodes with Epyc
2 episodes with Epyc
No more misinformation - AMD finally announced their Radeon RX 9060 XT, and it turns out that reports of a $449 MSRP were greatly exaggerated. And what the heck is going on with Ryzen and EPYC right now? We have EPYC branded Ryzen desktop chips, and Ryzen branded EPYC chips for workstations (gross oversimplification)?! Developers! (get fired), Silverstone retro, and also there's a clipboard.Timestamps:00:00 Intro01:45 Food with Josh03:57 Radeon RX 9060 XT announced, starts at 299 USD08:53 Threadripper 900012:07 Crucial's T710 Gen5 SSD announced13:54 Microsoft injecting "ai" into File Explorer21:36 A new dual-GPU graphics card in 2025??26:27 SilverStone adds FLP02 to retro case lineup29:14 MS brings back the DOS-era Edit program31:18 Netflix plans generative "ai" ads during streams in 202633:59 Podcast sponsor NordLayer35:29 (in)Security Corner53:49 Gaming Quick Hits59:00 Kent's NZXT H3 Flow mATX case review1:09:46 A quick look at MSI's RTX 5070 GAMING TRIO OC1:16:48 Picks of the Week ★ Support this podcast on Patreon ★
We were off last week, and somehow we STILL haven't reached Computex. So, another pre-show show, full of speculation (and speculative exploits). With no time to spare, we talk about Epyc, RX 9060 and RX 9070's, ceramic storage and Windows 3.1 - but not in that order.Timestamps: 00:00 Intro 00:56 Patreon 01:55 Food with Josh 03:58 Radeon RX 9060 XT clocks and pricing 05:54 RX 9070 GRE is a product that exists 07:14 WD_Black SN8100 is reaching for the Gen5 crown 09:59 AMD goes GRADO 18:19 Asetek patent expires - AiO competition will be fierce 20:22 Win 3.1 icons are hiding inside your Windows install RIGHT NOW 25:43 Microsoft fixes boot issues with Linux/Windows dual-boot systems 27:43 Podcast sponsor NordLayer 29:14 (in)Security Corner 44:08 Gaming Quick Hits 53:59 Jeremy has gone MAD (for ARGB) 1:03:28 Picks of the Week 1:17:59 Outro ★ Support this podcast on Patreon ★
AMDがエントリーサーバ向けCPU「EPYC 4005」を投入 Zen 5アーキテクチャベースでSocket AM5採用。 AMDは5月13日（米国太平洋夏時間）、エントリークラスサーバ向けの新型CPU「EPYC 4005シリーズ」を発表した。本CPUを搭載するサーバは、順次発売される予定だ。
Today in AI is a daily recap of the latest news and developments in the AI industry. See your story and want to be featured in an upcoming episode? Reach out at tonyphoang.com Nvidia CEO Jensen Huang and Cisco CEO Chuck Robbins recently highlighted the strategic importance of AI infrastructure in the global race for AI dominance. Their companies have partnered to integrate advanced AI technologies into enterprise systems, aiming to enhance operational efficiency, improve security, and drive innovation across various industries. This collaboration sets new standards for AI infrastructure, showcasing the critical role of AI in modern enterprise solutions. Under Lisa Su's leadership, AMD has transformed from a struggling company into a formidable competitor in the semiconductor industry. Significant advancements in high-performance computing, graphics technologies, and AI have been achieved, with products like the Ryzen and EPYC processors. Additionally, AMD has made notable strides in the AI GPU market with the Instinct MI325X chip and the open-source ROCm platform, further solidifying its position in the industry. ADQ and Energy Capital Partners have formed a $25 billion investment partnership to enhance power generation in the US, addressing the growing energy demands of data centers and other high-growth sectors. This initiative supports the rapid expansion of AI technologies and data centers while promoting renewable energy solutions. The partnership aims to meet the increasing electricity consumption in the United States, ensuring sustainable growth in the AI and data center industries.
Alex's server meets a tragic end—but its replacement is shaping up to be a beast. Meanwhile, Chris experiments with budget CO₂ detectors, sirens, and smart integrations. Plus some surprises!
Aujourd'hui, voici pourquoi Lisa Su, la PDG d'AMD, vient d'être nommée PDG de l'année 2024 par le magazine Time.Il faut dire qu'au milieu des annonces autour de la bataille de l'IA entre Qualcomm, Nvidia et Intel, AMD apparaît moins dans les titres de presse. Pourtant, la PDG de ce spécialiste de l'infrastructure vient d'être honorée.Alors pourquoi ce titre si prestigieux ?AMD a centuplé sa capitalisation boursièreEt bien pour commencer AMD a centuplé sa capitalisation boursière sous la direction de Lisa Su. En 2014, quand elle prend les commandes de ce géant des semi-conducteurs, l'entreprise traverse une période difficile.Mais grâce à une stratégie audacieuse et une vision claire, AMD est passé en 10 ans d'une capitalisation boursière de 2 milliards de dollars à plus de 200 milliards aujourd'hui.Un des plus grands tournants initié par la femme d'affaires a été le développement et la commercialisation de la gamme de processeurs AMD EPYC. Ce sont ces puces qui ont permis à AMD de devenir un acteur majeur dans les secteurs des serveurs informatiques et des centres de données. Sous le règne de Lisa Su, la part de marché de l'entreprise dans ce secteur est passée de 1 % à près de 34 %. De quoi damer le pion à l'éternel rival Intel, qui vient lui de perdre son PDG, mis de force à la retraite.Et les processeurs EPYC équipent aujourd'hui certains des superordinateurs les plus rapides et les plus économes en énergie au monde.La reine de l'innovationLe second point, c'est que Lisa Su est aussi une experte de l'innovation. Sous son leadership, AMD a investi massivement en recherche et développement, avec un montant de près de 6 milliards de dollars rien qu'en 2023.Ces investissements permettent à AMD de proposer désormais des solutions d'infrastructure pour l'intelligence artificielle, un domaine clé pour l'avenir.À titre d'exemple, AMD a récemment racheté Silo AI, un laboratoire d'IA en Europe, et ZT Systems, un fournisseur d'infrastructure spécialisé pour les géants du cloud.La plus grosse acquisition du secteur, c'est ellePour couronner le tout, Lisa Su a aussi marqué l'histoire des semi-conducteurs en réussissant la plus grosse acquisition jamais réalisée dans ce secteur. Il s'agit de celle de Xilinx, spécialiste de l'informatique adaptative.Une opération qui a fait d'AMD un leader incontournable du secteur.Née à Taïwan et diplômée du prestigieux Massachusetts Institute of Technology, Lisa Su a toujours été une pionnière. Elle a commencé sa carrière chez IBM et pilote aujourd'hui une entreprise à la pointe de l'innovation.Le ZD Tech est sur toutes les plateformes de podcast ! Abonnez-vous !Hébergé par Ausha. Visitez ausha.co/politique-de-confidentialite pour plus d'informations.
There's releases, bug fixes, and Windows news. We covered CAD, DigiKam, and Wine. Then new hardware and support, EPEL on WSL, and the Free Software Foundation's de-blobbed kenel. Then finally we cover the latest drama in the kernel and Code of Conduct enforcement. For tips we have pw-container, kdocker on Wayland, and Qucs-S. You can find the show notes at https://bit.ly/3ALweut and we'll see you next time! Host: Jonathan Bennett Co-Hosts: Jeff Massie and Ken McDonald Want access to the video version and exclusive features? Become a member of Club TWiT today! https://twit.tv/clubtwit Club TWiT members can discuss this episode and leave feedback in the Club TWiT Discord.
It's Six Five Media at AMD Advancing AI, with our hosts Daniel Newman and Patrick Moorhead joined by AMD's SVP of Artificial Intelligence Vamsi Boppana, and Brad McCredie SVP of Data Center GPU and Accelerated Processing, to discuss the chipmaker's full-stack data center solutions, EPYC processor market share, and the evolving roles of GPUs and CPUs in AI. Their discussion covers: The evolution of AMD's strategy from providing compute engines to offering comprehensive data center solutions The integration of AI, Instinct Accelerators, networking compute, and EPYC CPUs into full stack solutions amid changing customer preferences The significant market share gain by AMD's EPYC processors, customer feedback, and how AMD continues to execute its roadmap effectively The emerging landscape of AI, with a focus on the balance between GPUs and CPUs in advancing AI technologies and what customers should consider as they venture into AI
George Cozma of Chips and Cheese joined Bryan, Adam, and the Oxide Friends to talk about AMD's new 5th generation EPYC processor, codename: Turin. What's new in Turin and how is Oxide's Turin-based platform coming along?In addition to Bryan Cantrill and Adam Leventhal, we were joined by special guest George Cozma, as well as Oxide colleagues Robert Mustacchi, Eric Aasen, Nathanael Huffman, and the quietly observant Aaron Hartwig.Some of the topics we hit on, in the order that we hit them:Chips and Cheese: AMD's Turin: 5th Gen EPYC LaunchedEnd of the Road: An Anandtech FarewellCentaur TechnologyAVX-512Zen5's AVX512 Teardown + More...Thermal Power Design (TDP)OxF: Rack Scale Networking (use of p4)P4AGESAOxF: The Network Behind the Network (Oxide server recovery)openSILphoronix: openSILPCB backdrillingOxF: AMD's MI300 (APUs)dtrace.conf(24) -- The DTrace unconference, December 11th, 2024If we got something wrong or missed something, please file a PR! Our next show will likely be on Monday at 5p Pacific Time on our Discord server; stay tuned to our Mastodon feeds for details, or subscribe to this calendar. We'd love to have you join us, as we always love to hear from new speakers!
Six Five Media hosts Daniel Newman and Patrick Moorhead are at AMD Advancing AI where they're joined by AMD's Forrest Norrod, Executive Vice President and General Manager, Data Center Solutions Business Group, for a conversation on AMD's strategic evolution toward becoming a data center solutions provider and the latest collaboration driving forward AI and chip ecosystems. Their discussion covers: An overview of AMD's evolved strategy from compute engines to comprehensive data center solutions AMD's integration of AI, Instinct Accelerators, and EPYC CPUs, and the role of ecosystem collaboration in offering full-stack solutions The significant impact of EPYC in the data center market, AMD's market share growth, and customer feedback The evolving landscape of AI technologies, the role of GPUs and CPUs, and strategic advice for customers exploring AI
Welcome to this week's edition of “MI&S Datacenter Podcast” I'm Patrick Moorhead with Moor Insights & Strategy, and I am joined by co-hosts Matt, Will, and Paul. We analyze the week's top datacenter and datacenter edge news. We talk compute, cloud, security, storage, networking, operations, data management, AI, and more! AMD Is Bringing Sexy Back To Networking https://x.com/WillTownTech/status/1844465726226301209 OpenAI o1 Is PhD Smart https://openai.com/index/learning-to-reason-with-llms/ Mo' Cores Mo' Cache - AMD PI https://www.amd.com/en/products/processors/server/epyc/9005-series.html Qualcomm Gets Edgy With Campus & Branch Connectivity Infrastructure https://x.com/WillTownTech/status/1843375274781749430 Europe Gets A Quantum Data Center https://www.ibm.com/quantum/blog/europe-quantum-datacenter-software Mind The (AI) Gap - AMD PII https://www.amd.com/en/products/accelerators/instinct/mi300.html Disclaimer: This show is for information and entertainment purposes only. While we will discuss publicly traded companies on this show. The contents of this show should not be taken as investment advice.
Our big story this week is from AMD. They're opening up their wallet to the tune of nearly $5 billion to buy ZT Systems. The two companies have had a preexisting partnership, with ZT having collaborated on the EPYC processor lines. ZT's biggest customers are AWS and Azure, as ZT specializes in hyperscale AI systems that are bought by the rack. This move follows a very recent acquisition of Silo AI, which we covered on the Rundown, as well as their last big acquisition of Xilinx. There's a lot to unpack here and the Futurum Group has had some amazing coverage of this deal so far. Stephen, let's start with you. What does ZT Systems have that makes them so attractive to AMD. Time Stamps: 0:00 - Welcome to the Rundown 1:48 - Morpheus Data Acquired by HPE 5:31 - Launchable Acquired by CloudBees 8:26 - Kioxia Reveals Broadband SSD 12:43 - DigiCert to Acquire Vercara to Expand Security Portfolio 16:25 - Western Digital Races Past NetApp with All-Flash OpenFlex 20:11 - Massive Data Leak From Plaintext Passwords 24:34 - ZT Systems to be Acquired by AMD 40:15 - The Weeks Ahead 42:19 - Thanks for Watching Hosts: Tom Hollingsworth: https://www.twitter.com/NetworkingNerd Stephen Foskett: https://www.twitter.com/SFoskett Follow Gestalt IT Website: https://www.GestaltIT.com/ Twitter: https://www.twitter.com/GestaltIT LinkedIn: https://www.linkedin.com/company/Gestalt-IT Tags: #Rundown, #CyberSecurity, #EPYC, @AMD, @ZTSystems, @WesternDigital, @NetApp, @DigiCert, @Vercara, @Kioxia, @Launchable, @CloudBees, @MorpheusData, @HPE, @GestaltIT, @TechFieldDay, @TheFuturmGroup, @Sfoskett, @NetworkingNerd,
Lion Cove is only getting a 14% IPC increase, and we discuss all Computex 2024 News! [SPON: Go to Jawa.gg to build your next PC Hassle Free: https://jawa.link/MLIDMay24 ] [SPON: New Customers get $25 OFF $100 Orders at Micro Center! https://micro.center/1gje ] [SPON: $23 Win11 Pro w/ "brokensilicon“ at CDKeyOffer: https://www.cdkeyoffer.com/cko/Moore11 ] [SPON: Shop Micro Center's Top Deals for May: https://micro.center/iy37 ] [SPON: Get the latest Tech News, Reviews, and Guides at MC News: https://micro.center/246240 ] 0:00 Tom needs Sleep, but he saw Modest Mouse! (Intro Banter) 4:15 Onshoring Silicon Misconceptions (Corrections) 10:01 Midrange GPU Pricing's INSANE, AMD Rebaits Backfires 20:10 Should you sell your 4090? How strong can RDNA 4 be? 26:23 Nvidia Unveils Vera Rubin & Snubs Gamers at Computex 35:20 Is Nvidia overhyped? Is the Copilot+ TOPs requirement arbitrary? 37:41 AMD Reveals EPYC Turin, MI325X, MI350X, and more! 47:18 Zen 5 Architecture & Ryzen 9000 Granite Ridge Detailed! 58:23 Should you wait for B850 or buy a 7800X3D w/ B650 now? 1:02:54 AMD AI 9 300 Strix Unveiled! 1:09:15 Does AMD know how to be #1? What's the point of Sound Wave? 1:18:08 Qualcomm Snapdragon X Elite gains TONS of OEM Support 1:26:26 Intel Lunar Lake Analysis – Red Flags Galore! 1:40:55 5090 28GB, 17.6 Gbps DDR6, EPYC 4004, GT 710 GDDR6 (Wrap-up) 1:43:47 Economic Bubbles, Public Perception of AI, Hot Chips Sponsors (Final RM) https://youtu.be/o_qlPl58qcY?si=Cy4Xy2yWifd_42d8 https://amzn.to/4eb1Qsj https://www.youtube.com/live/pKXDVsWZmUU?si=qmpygSerkontpjAd https://blogs.nvidia.com/blog/computex-2024-jensen-huang/ https://www.nvidia.com/en-us/geforce/news/g-assist-ai-assistant/ https://videocardz.com/newz/nvidia-unveils-vera-rubin-as-next-gen-gpu-architecture-patron-rubin-ultra-to-sport-12-hbm4-stacks https://www.youtube.com/live/MCi8jgALPYA?si=eJ6SRFpx8ExKPxBW https://www.amd.com/en/newsroom/press-releases/2024-6-2-amd-accelerates-pace-of-data-center-ai-innovation-.html https://www.amd.com/en/newsroom/press-releases/2024-6-2-amd-unveils-next-gen-zen-5-ryzen-processors-to-p.html https://www.amd.com/en/newsroom/press-releases/2024-6-2-amd-unveils-next-gen-zen-5-ryzen-processors-to-p.html https://www.youtube.com/live/5R3QpKMciEw?si=EqoIImjaJ13rhFn0 https://www.qualcomm.com/news/releases/2024/06/qualcomm-computex-2024-keynote-unveils--the-pc-reborn--with-snap https://signal65.com/wp-content/uploads/2024/05/NewSurfaceLaptop2024_Signal65LabInsights_r1.01.pdf https://www.intel.com/content/www/us/en/events/computex.html https://www.anandtech.com/show/21425/intel-lunar-lake-architecture-deep-dive-lion-cove-xe2-and-npu4/6 https://twitter.com/thraxbert/status/1797809739138728240 https://videocardz.com/newz/nvidia-rtx-5090-new-rumored-specs-28gb-gddr7-and-448-bit-bus https://videocardz.com/newz/next-gen-lpddr6-memory-to-hit-up-to-14-4-gbps-data-rate-ddr6-up-to-17-6-gbps https://videocardz.com/newz/amd-launches-epyc-4004-cpus-on-am5-package-up-to-16-zen4-cores-3d-v-cache-for-699 https://www.techpowerup.com/gpu-specs/radeon-rx-6800.c3713 https://www.microcenter.com/product/674503/amd-ryzen-7-7800x3d-raphael-am5-42ghz-8-core-boxed-processor-heatsink-not-included https://youtu.be/YywHyv66If4
AMDがデータセンター向けCPU「第5世代EPYC」の登場を予告 Zen 5アーキテクチャ採用で最大128コア 2024年後半に発売予定。 AMDは6月3日、Zen 5アーキテクチャを適用したデータセンター／サーバ向け新型CPU「第5世代EPYCプロセッサ」（開発コード名：Turin）を2024年後半に投入することを発表した。
AMDが中小規模サーバ向けCPU「EPYC 4004シリーズ」を発表 Socket AM5採用でコストを抑制。 AMDは5月20日（米国太平洋夏時間）、PCベースの中小規模サーバ向けの新型CPU「EPYC 4004シリーズ」を発表した。本CPUを搭載するサーバは、順次発売される予定だ。
AMD set to release Epyc 4004 on AM5; China's Zhaoxin CPU matches Skylake
Alex's new Epyc server build, and Jon Seager from Canonical joins us to chat about Nix in the homelab, packaging Scrutiny, and how Nix fits with existing infrastructure management tools. Special Guest: Jon Seager .
Alex's new Epyc server build, and Jon Seager from Canonical joins us to chat about Nix in the homelab, packaging Scrutiny, and how Nix fits with existing infrastructure management tools. Special Guest: Jon Seager .
This week Jeff is accompanied by Steve (the Devil's Chicken, himself) and they are discussing everything happening in the world of tech, computers, gaming, craft beer and cocktails.
Lenovo has been famous for their laptops for a long time, but they also have a strong business in computer servers. They offer hardware for data centers as well as edge locations, along with a SaaS style pricing model called TruScale. According to Robert Daigle, Director of Global AI at Lenovo, a new system called the Lenovo ThinkEdge SE455 V3 makes it possible to run compute-heavy tasks such as generative AI in a workplace at "the edge." The SE455 is only half as loud as comparable servers, and can survive at a temperature of 55°C, and therefore can co-exist with people in a typical office or research center. The SE455 is powerful, with 64 cores (AMD 8004 EPYC processors), and energy-efficient. Thus, compute-heavy applications that most organizations used to run in the cloud can be done on-premises. Learn more about Lenovo: https://www.lenovo.com/us/en/servers-storage/solutions/analytics-ai/ Health IT Community: https://www.healthcareittoday.com/
AMD kehren endlich mit neuen Threadripper (Pro) zurück in den HEDT-Markt: Bis zu 96-Zen-4-Kerne in einem Paket (Pro) und verdammt viele PCIe-Lanes. Intels 14. Generation für den Consumer-Bereich nicht nur im Vergleich langweilig, sondern einfach generell. Einer der belanglosesten Releases seit langem; eigentlich nur 200 MHz mehr, außer beim 14700K(F): Der bringt vier E-Cores mehr als der 13700K(F). Spannendes passiert in China mit dem Prozessor FTC870 von Phytium, der laut Spec-CPU2017-Benchmark (fragwürdiger Benchmark) mit AMD Epyc 7443 (Zen 3) konkurrieren möchte. Und Qualcomm möchte zusammen mit Google einen SoC auf Basis von RISC-V zur Serienreife bringen und so eine Alternative zu Arm haben. Nach dem Nvidia-Debakel und dem Börsengang von Arm sehr begrüßenswert. Wir haben diesmal viel zu Autos (yay): BMW und Mercedes schielen neidisch auf Tesla oder Nio und möchten auch Direktvertrieb, um die Preise zu kontrollieren und Händler zu umgehen. Alles zum Wohle des Kunden natürlich. Und dann kommt da Citroën mit dem neuen "Kleinwagen" Ë-C3. Preislich ganz interessant für ein E-Auto, aber "Kleinwagen"? Naja. Mike hat sich das "The Bloodline" im Early Access angesehen, ein Indie-Sandbox-RPG mit Anleihen an Elder Scrolls, Mount & Blade und Survival Games. Interessante Ansätze, viel Potential, aber noch sehr rumpelig. Das braucht noch viel Zeit und viele Patches. Zum Schluss haben wir leuchtende Petunien und Spinnen-Geschichten. Viel Spaß mit Folge 175! Sprecher: Meep, Mohammed Ali Dad, Michael KisterProduktion: Michael KisterTitelbild: Mohammed Ali DadBildquellen:Aufnahmedatum: 20.10.2023 Besucht unsim Discord https://discord.gg/SneNarVCBMauf Twitter https://twitter.com/technikquatschauf Bluesky https://bsky.app/profile/technikquatsch.bsky.socialauf Youtube https://www.youtube.com/@technikquatsch 00:00:00 Eigenpromo zum Switch-2-Special mit Mats, Unholy (Wars)https://technikquatsch.de/special-die-switch-2-technikquatsch-fantasiert-feat-special-guest-mats/, auch auf Youtube: https://www.youtube.com/watch?v=oKEslNJz5HsWitchery - Unholy Wars https://www.youtube.com/watch?v=zt9j9mgNSug 00:11:23 AMD Threadripper 7000 (Pro)https://www.anandtech.com/show/21092/amd-unveils-ryzen-threadripper-7000-family-zen-4-for-workstations-and-hedthttps://www.computerbase.de/2023-10/amd-ryzen-threadripper-pro-7000-mit-64-zen-4-kernen-als-hedt-cpu-auch-fuer-den-desktop/ 00:18:56 Intel 14. Generation: 14600K(F), 14700K(F), 14900K(F)https://www.computerbase.de/2023-10/intel-core-i9-14900k-i7-14700k-i5-14600k-test/ 00:24:34 CPU-Hersteller Phytium aus China möchte mit Zen 3 Epyc konkurrierenhttps://www.golem.de/news/phytium-ftc870-chinesische-cpu-konkurriert-mit-amd-zen-3-2310-178657.html 00:29:07 Qualcomm stellt SoC auf RISC-V-Basis für WearOS vorhttps://www.computerbase.de/2023-10/qualcomm-und-google-smartwatch-soc-mit-risc-v-cpu-fuer-wear-os-geplant/https://www.qualcomm.com/news/releases/2023/10/qualcomm-to-bring-risc-v-based-wearable-platform-to--wear-os-by- 00:36:26 BMW und Mercedes wollen auf Direktvertrieb setzenhttps://www.handelsblatt.com/unternehmen/industrie/autoindustrie-bmw-plant-direktvertrieb-und-will-die-preise-kontrollieren-/29432448.html 01:00:01 der neue "Kleinwagen" Citroën Ë-C3https://www.golem.de/news/citroen-e-c3-von-den-schwierigkeiten-einen-e-kleinwagen-zu-bauen-2310-178661.html 01:20:01 "The Bloodline" Early Access Preview: interessantes Sandbox RPG mit Potential, das noch reifen musshttps://store.steampowered.com/app/1159290/The_Bloodline/ 01:32:56 "glow in the dark" Petunien in den USA zum Verkauf zugelassenhttps://www.heise.de/news/Gruenes-Licht-Zulassung-fuer-den-Verkauf-von-leuchtenden-Petunien-9337678.html 01:41:40 CW: Spinnen 01:48:06 Terrifier auf Amazon Prime als geschnittene FSK 18https://www.amazon.de/Terrifier-David-Howard-Thornton/dp/B07KRF9XYKhttps://www.schnittberichte.com/schnittbericht.php?ID=71543 01:50:45 Ende
AMD ha anunciado la disponibilidad de los nuevos Procesadores AMD EPYC Serie 8004, que están diseñados para ofrecer eficiencia energética y rendimiento en cargas de trabajo. Estos procesadores, con el núcleo "Zen 4c", son ideales para infraestructuras con limitaciones de espacio y potencia, cubriendo desde el borde inteligente hasta el centro de datos.
Welcome back to Rich in Real Life! Today I am excited to share a incredible intimate conversation with a dear friend of mine, Megan, who's a business systems expert but also my best friend, and my business partner! In this episode we dove deep into the importance of systems in both life and business - We talked about how systems can help you grow, how to avoid the shiny object syndrome, and ways to create an exceptional client experience in your business. Megan also shared her personal journey of realizing how she needed to work on herself, which eventually led her to the realization that many business challenges are not just system problems, but sometimes therapy problems too! We also talked about the inevitable messiness that comes with building and refining systems and how crucial it is to embrace outsourcing to free up your time. So, whether you're a business owner on the fence about systems or someone just looking to simplify and thrive in life, this episode is sure to offer some valuable insights into how you can truly become Rich in Real Life! KEY POINTS - How embracing systems can drive growth, reduce distractions, and enhance client experience - The importance of differentiating between system problems and personal growth challenges in business - The messy but vital process of building and refining systems for business success - How outsourcing tasks will save time and provide you freedom to focus on income-generating activities - The foundation for success in business is found in your personal life QUOTABLES "It's really hard for people to understand that transactions aren't a dollar sign." - Jessica Hurley. "Go ahead and take a look at your life. How can you make it more simple?" - Megan. "I know everyone's talking about outsourcing, but really consider writing down how much it is to outsource certain things." - Jessica Hurley. RESOURCES Megan Galane Instagram | @megangalane LinkedIn | @megangalane Love what you're hearing? Follow Jessica Hurley and share the love! IG | @jessicahurley__ Rich In Real Life is produced by EPYC Media Network
Dans cet épisode de rentrée, Antonio et Arnaud ont le plaisir d'accueillir Katia Aresti dans l'équipe. Ils passent en revue les dernières nouveautés et sujets chauds de cette rentrée, notamment la sortie de Java 21, les nouvelles versions de Quarkus, Micronaut, Hibernate, NodeJS, Redis, et bien d'autres encore. Ils discutent de sujets plus généraux tels que l'observabilité, la nouvelle tendance “Platform Engineering”, et la productivité des développeurs. Ils abordent aussi les sujets sur la sécurité, tels que les failles sur les CPUs Intel et AMD, ainsi que la vie privée, avec les Tracking APIs de Chrome, Firefox et le projet de loi SREN. Le tout est agrémenté de sa dose d'IA, avec des librairies telles que Semantic Kernel, ainsi que des sujets plus haut niveau tels que Google Gemini, Meta GPT, LLama 2, et les biais et la consommation énergétique de l'IA. Enregistré le 8 septembre 2023 Téléchargement de l'épisode LesCastCodeurs-Episode–299.mp3 News Langages Apache Groovy a 20 ans! https://twitter.com/ApacheGroovy/status/1695388098950217909 L'annonce du lancement du projet par James Strachan https://web.archive.org/web/20030901064404/http://radio.weblogs.com/0112098/2003/08/29.html Le projet a depuis énormément évolué et après plusieurs vies a été adopté par la fondation Apache en 2015 Java 21 arrive le 19 septembre https://www.infoworld.com/article/3689880/jdk–21-the-new-features-in-java–21.html. C'est la nouvelle LTS Pas mal de nouvelles fonctionnalités comme les virtual threads, le pattern matching sur les switch, sequenced collections … Retrouvez le 19 septembre une interview de Jean-Michel Doudoux par Charles Sabourdin pour l'épisode 300 des castcodeurs! Librairies Semantic Kernel pour Java est (en train de) sorti: https://devblogs.microsoft.com/semantic-kernel/introducing-semantic-kernel-for-java/ Framework OSS pour faire de l'IA .Net et Python Java 0.2.7 Alpha est publié Kernel car il est tout petit Se connecte à plusieurs fournisseurs (aujourd'hui OpenAI, Azure AI, Hugging Face), plusieurs DB vectorielles, plusieurs template de prompt (suit la specification de OpenAI) OpenSSL qui committe https://www.openssl.org/blog/blog/2023/07/17/who-writes-openssl/ en majorité des OSS payés puis des gens payés par leur boite et enfi des contributeurs non payés c'est ne passant rapide mais ca montre que depuis heartbleed, ca a changé Micronaut 4.1.0 https://micronaut.io/2023/09/01/micronaut-framework–4–1–0-released/ Bean Mappers pour créer automatiquement une correspondance entre un type et un autre un Introspection Builder l'annotation @Introspected pour générer un builder dynamique si un type ne peut être construit que via un modèle builder améliorations pour les développeurs utilisant Kotlin Symbol Processing (KSP) Quarkus 3.3.1 / 3.3.2 https://quarkus.io/blog/quarkus–3–3–1-released/ https://quarkus.io/blog/quarkus–3–3–2-released/ Pas mal de fixes https://github.com/quarkusio/quarkus/releases/tag/3.3.1 https://github.com/quarkusio/quarkus/releases/tag/3.3.2 Il est important de noter qu'un problème de dégradation des performances et de la mémoire a été introduit dans Quarkus 3.3. Ce problème est corrigé dans Quarkus 3.3.2. Hibernate ORM 6.3.0 et 6.2.8 https://hibernate.org/orm/ et Hibernate Reactive 2.0.5 un support initial de la spécification Jakarta Persistence 3.2 Un nouveau guide d'introduction Hibernate 6, un nouveau guide de syntaxe et de fonctionnalités pour le langage de requête Hibernate (Hibernate Query Language) Annotation @Find sur des méthodes -> créer des méthodes de recherche similaires aux méthodes de requête Reactive compatible avec Hibernate ORM 6.2.8.Final, certains changements d'api Infrastructure Une série d'articles sur l'observabilité par Mathieu Corbin Observability: tout ce que vous avez toujours voulu savoir sur les métriques: https://www.mcorbin.fr/posts/2023–07–04-metriques/ Tracing avec Opentelemetry: pourquoi c'est le futur (et pourquoi ça remplacera les logs): https://www.mcorbin.fr/posts/2023–08–20-traces/ L'auteur reprend les bases sur l'observabilité. Qu'est ce qu'une métrique ? Les labels, les cardinalités Les types de métriques (Compteurs, jauges, quantiles et histogrammes) C'est quoi le tracing ? Traces, Spans, Resources, Scopes qu'est ce que c'est? Les Events pour remplacer les logs? Web NodeJS 20.6.0 est disponible et ajoute le support des fichiers .env https://philna.sh/blog/2023/09/05/nodejs-supports-dotenv/ Configurable avec l'option --env-file Le fichier .env peut contenir des variables d'environnement et commentaires # Attention par contre: pas de lignes multiples ni d'extension de variables Vous pouvez par exemple configurer NODE_OPTIONS avec ce système Data Redis 7.2 est sorti ! https://redis.com/blog/introducing-redis–7–2/ Auto-tiering : cette nouvelle fonctionnalité permet de stocker les données sur des supports de stockage différents, en fonction de leur importance et de leur fréquence d'accès. Cela permet d'améliorer les performances et la scalabilité de Redis. RESP3 : cette nouvelle version du protocole RESP permet une communication plus efficace entre Redis et les clients. Improvements to performance : de nombreuses améliorations de performances ont été apportées à Redis 7.2, notamment pour les opérations de lecture et d'écriture. New commands : plusieurs nouvelles commandes ont été ajoutées à Redis 7.2, notamment : CLIENT NO-TOUCH : cette commande permet d'empêcher un client d'être touché par une opération AOF ou RDB. WAITAOF : cette commande permet d'attendre que l'AOF soit écrite avant de poursuivre l'exécution. Dans le podcast sont cités les hot replacement des Redis, comme https://www.dragonflydb.io/ Architecture Article sur Google Gemini et sa capacité a battre ChatGPT https://www.semianalysis.com/p/google-gemini-eats-the-world-gemini Google a raté les premiers pas (ils avient le meilleur LLM public avant ChatGPT 3) ET les chercheurs qui invente le champs des LLMs Google va 5x ChatGPT–4 avant al fin de l'année, mais vont-il les publier les chercheurs se tirent la bourre sur le nombre de GPU (H100) auxquels ils ont accès ; ce sont lers grosses orga comme Meta OpenAI Google et les autres qui lutent avec des GPU qui n'ont pas assez de VRAM et ce qu'ils vont faire c'est de la merde et sans consequence le peuple utilise le modele dense de LLAMA mais pour les environnements contraints ca serait mieux des sparse models et du speculative decoding. ils devraient se concentre sur la performance de modele qui utilise plus de compute et memoire en evitant de consommer de la bande passante de memoire, c'est ce que l'edge a besoin les benchmarks public ne mesurent pas des choses utiles meme hugging faces est dans la category des pauvres de GPU Nvidia est entrain de se construire une machine de guerre (service) la chine et les us vont etre en competition mais l'europe qui fait du GPU pauvre ne va pas s'en sortir les startups ne peuvent pas payer les GPU en actiosn, il faut du cash Tout le monde rempli les poches de NVidia, sand Google Gogole grossi exponentiellement ses propres GPUs Meta GPT https://www.infoq.com/news/2023/08/metagpt-agent-collaboration/ IA: les biais et énergie qui consomme par Leslie Miley tech advisor du CTO de Microsoft https://www.infoq.com/presentations/ai-bias-sustainability nouvels infranstructures consommation énergétique et d'eau des data center pour IA est terriblement coûteuse l'impact des infrastructures sur les comunautés (bruit) explique bien son point de vu sur les problèmes d'amplification des biais du IA propose des stratégies pour mitiger l'impact negatif Kubeflow toolkit pour deployer machine learning (ML) workflow en Kubernetes est accepté par la CNCF (Cloud Native Computing Foundation) https://www.infoq.com/news/2023/08/kubeflow-cncf-project Méthodologies Measuring developer productivity? A response to McKinsey by Kent Beck and Gergely Orosz (pragmaticengineer.com) https://tidyfirst.substack.com/p/measuring-developer-productivity McKinsey a sorti un article où ils expliquent la recette miracle recherchée par tous les managers comme le graal: Comment mesurer la productivité des développeurs? (faut bien vendre du conseil) Kent et Gergely partent d'un model mental de description de la création de valeur par le développeur pour ensuite voir quels sont les besoins de mesurer la productivité et comparent cela avec d'autres secteurs (la vente, le support, le recrutement). Ils concluent cette première partie avec les compromis à faire pour que ce type de mesures ait un intérêt sans impacter trop négativement les développeurs un autre article dans la même lignée de Martin Fowler https://martinfowler.com/bliki/CannotMeasureProductivity.html Et si on parlait de Platform Engineering ? DevOps vs. SRE vs. Platform Engineering (humanitec.com) What is platform engineering? (gartner.com) / What is platform engineering? (platformengineering.org) Internal Developer Platform Cognitive load Team topologies Engineering Effectiveness (thoughtworks.com) and Maximize your tech investments with Engineering Effectiveness (thoughtworks.com) Ces différents articles retracent la génèse du concept de Platform Engineering L'activité de Platform Engineering vient en réponse à la charge cognitive rajoutée aux équipes techs dans des transitions DevOps loupées (You build it, you run it … et vous vous débrouillez). Cela conduit à la création de golden paths et d'une Internal Developers Platform qui doit proposer en interne les services nécessaires aux équipes pour livrer leurs produits le lus efficacement possible tout en suivant les critères de qualité, de compliance de l'entreprise. Pour en savoir plus, une table ronde à laquelle Arnaud a participé en Juillet : https://youtu.be/N-tN7HUA4No?si=2P0wSqG32MLWUlGq On call Process (Astreinte) , startup TinyBird par VP Engineering Félix López (ex google, ex eventbrite) https://thenewstack.io/keeping-the-lights-on-the-on-call-process-that-works/ Si votre produit est SAAS, on doit avoir des astreintes. Cela impose un lourd fardeau à ceux qui doivent être en astreinte,, surtout en petite entreprise Petites entreprises évitent avoir un processus d'astreinte formel pour éviter le stress. Cela crée dans la pratique plus de stress: Si personne n'est responsable, tout le monde est responsable. Tinybird est la plateforme de données en temps réel pour les développeurs et les équipes de données. Pré création du process formel chez Tinybird: désorganisé, non structuré et stressant Mise en place: Principes fondamentaux d'un processus d'astreinte: L'astreinte n'est pas obligatoire, minimiser le bruit, pas seulement pour les SRE, alert = runbook, avoir des backups pour la personne en astreinte, appeler quelqu'un devrait être la dernière solution, minimiser le temps en astreinte L'article explique comment ils sont passé regarder chaque alerte (comprehensible?, exploitable?), puis avoir un board grafana pour chacune et plan spécifique. Une fois le tri fait, tout migré vers un seul channel de com, et manuel d'astreinte pour chaque alerte. Itérer. Multiples benefices sur le long terme: rapports d'incident ouvert, atténuer les problèmes futurs, renforcement la propriété et les connaissances du code et systèmes au sein de toute l'équipe etc. Sécurité Downfall, une nouvelle faille de sécurité sur les processeurs intel ( https://www.lemondeinformatique.fr/actualites/lire-la-faille-downfall-met-a-mal-des-milliards-de-processeurs-intel–91247.html ) et AMD ne fait pas mieux avec une faille nommée Inception (https://www.lemondeinformatique.fr/actualites/lire-les-puces-amd-vulnerables-a-la-faille-inception–91273.html) Downfall, La vulnérabilité est due à des fonctions d'optimisation de la mémoire dans les processeurs Intel qui révèlent involontairement les registres matériels internes aux logiciels. Cela permet à des logiciels non-fiables d'accéder à des données stockées par d'autres programmes, qui ne devraient normalement pas être accessibles. Tous les PC ou ordinateurs portables équipés de processeurs Intel Core de la 6e génération Skylake jusqu'aux puces Tiger Lake de 11e génération incluses contiennent cette faille. Les derniers processeurs Core 12e et 13e génération d'Intel ne sont pas concernés. Inception, nécessite un accès local au système pour être potentiellement exploité ce qui en limite de fait la portée. Tous les processeurs AMD depuis 2017 sont touchés, incluant les derniers modèles Zen 4 Epyc et Ryzen Comment désactiver le nouveau tracking publicitaire ciblé sur Chrome https://www.blogdumoderateur.com/chrome-comment-desactiver-tracking-publicitaire-cible/ Google a annoncé en juillet le déploiement de sa nouvelle API Topics, permettant « à un navigateur de partager des informations avec des tiers sur les intérêts d'un utilisateur tout en préservant la confidentialité ». C'est cette API, incluse dans la version Chrome 115 de juillet 2023, qui est censée remplacer les cookies tiers. Loi, société et organisation Une nouvelle definition d'open pour Llama 2? https://opensourceconnections.com/blog/2023/07/19/is-llama–2-open-source-no-and-perhaps-we-need-a-new-definition-of-open/ c'est relativement “open” mais il y a des restrictions donc pas open source pas plus de 700 M d'utilisateurs par mois pas le droit d'utiliser Llama pour améliorer d'autres modèles autres que dse dérivés de Llama et c'est le modele final qui est ouvert, pas la sauce pour le construire, donc pas de maven build ni le “source code” pour y arriver “from scratch” attention au risuqe de sacrivier open source pour avoir l'IA plus vite, plus facile HashiCorp passe tous ses projets open source en BSL, comme Confluent, Mongo, Redis, Elastic, etc https://thenewstack.io/hashicorp-abandons-open-source-for-business-source-license/ Couverture par InfoQ https://www.infoq.com/news/2023/08/hashicorp-adopts-bsl/ Fork de Terraform : OpenTF, avec pour objectif de rejoindre la CNCF https://opentf.org/announcement Stack overflow annonce Overflow AI https://www.infoq.com/news/2023/09/stackoverflow-overflowai/ l'intégration de l'IA générative dans leur plateforme publique, Stack Overflow for Teams, ainsi que de nouveaux domaines de produits IA/ML aident à générer des balises initiales et à suggérer des paires question-réponse, permettant aux développeurs de se concentrer sur l'amélioration et la précision Amélioration des Capacités de Recherche Les forums de questions-réponses basés sur la communauté sont le cœur battant de Stack Overflow. Selon Prashanth Chandrasekar, PDG de Stack Overflow, l'objectif d'OverflowAI est d'améliorer la communauté de diverses manières plutôt que de la remplacer complètement. Vous avez entendu parler du projet de loi SREN ? http://share.mozilla.org/817319645t Le gouvernement français prépare une loi qui pourrait menacer la liberté sur Internet. Le projet de loi visant à sécuriser et réguler l'espace numérique (SREN) obligerait les navigateurs web, comme Mozilla Firefox, à bloquer des sites web directement au niveau du navigateur. Mozilla lance une pétition pour retirer cette n-ieme solution stupide pour censurer Internet Conférences La liste des conférences provenant de Developers Conferences Agenda/List par Aurélie Vache et contributeurs : 8 septembre 2023 : JUG Summer Camp - La Rochelle (France) 14 septembre 2023 : Cloud Sud - Toulouse (France) & Online 18 septembre 2023 : Agile Tour Montpellier - Montpellier (France) 19 septembre 2023 : Salon de la Data Nantes - Nantes (France) & Online 19–20 septembre 2023 : Agile en Seine - Paris (France) 21–22 septembre 2023 : API Platform Conference - Lille (France) & Online 22 septembre 2023 : Agile Tour Sophia Antipolis - Valbonne (France) 25–26 septembre 2023 : BIG DATA & AI PARIS 2023 - Paris (France) 28–30 septembre 2023 : Paris Web - Paris (France) 2–6 octobre 2023 : Devoxx Belgium - Antwerp (Belgium) 6 octobre 2023 : DevFest Perros-Guirec - Perros-Guirec (France) 10 octobre 2023 : ParisTestConf - Paris (France) 11–13 octobre 2023 : Devoxx Morocco - Agadir (Morocco) 12 octobre 2023 : Cloud Nord - Lille (France) 12–13 octobre 2023 : Volcamp 2023 - Clermont-Ferrand (France) 12–13 octobre 2023 : Forum PHP 2023 - Marne-la-Vallée (France) 13–14 octobre 2023 : SecSea 2K23 - La Ciotat (France) 17–20 octobre 2023 : DrupalCon Lille - Lille (France) 19–20 octobre 2023 : DevFest Nantes - Nantes (France) 19–20 octobre 2023 : Agile Tour Rennes - Rennes (France) 26 octobre 2023 : Codeurs en Seine - Rouen (France) 30 septembre 2023 : ScalaIO - Paris (France) 26–27 octobre 2023 : Agile Tour Bordeaux - Bordeaux (France) 26–29 octobre 2023 : SoCraTes-FR - Orange (France) 10 novembre 2023 : BDX I/O - Bordeaux (France) 15 novembre 2023 : DevFest Strasbourg - Strasbourg (France) 16 novembre 2023 : DevFest Toulouse - Toulouse (France) 18–19 novembre 2023 : Capitole du Libre - Toulouse (France) 23 novembre 2023 : DevOps D-Day #8 - Marseille (France) 23 novembre 2023 : Agile Grenoble - Grenoble (France) 30 novembre 2023 : PrestaShop Developer Conference - Paris (France) 30 novembre 2023 : WHO run the Tech - Rennes (France) 6–7 décembre 2023 : Open Source Experience - Paris (France) 7 décembre 2023 : Agile Tour Aix-Marseille - Gardanne (France) 7–8 décembre 2023 : TechRocks Summit - Paris (France) 8 décembre 2023 : DevFest Dijon - Dijon (France) 31 janvier 2024–3 février 2024 : SnowCamp - Grenoble (France) 6–7 mars 2024 : FlowCon 2024 - Paris (France) 19–22 mars 2024 : KubeCon + CloudNativeCon Europe 2024 - Paris (France) 28–29 mars 2024 : SymfonyLive Paris 2024 - Paris (France) 17–19 avril 2024 : Devoxx France - Paris (France) 25–26 avril 2024 : MiXiT - Lyon (France) 25–26 avril 2024 : Android Makers - Paris (France) 6–7 juin 2024 : DevFest Lille - Lille (France) Nous contacter Pour réagir à cet épisode, venez discuter sur le groupe Google https://groups.google.com/group/lescastcodeurs Contactez-nous via twitter https://twitter.com/lescastcodeurs Faire un crowdcast ou une crowdquestion Soutenez Les Cast Codeurs sur Patreon https://www.patreon.com/LesCastCodeurs Tous les épisodes et toutes les infos sur https://lescastcodeurs.com/
Google Next Eve! Welcome episode 225 of The CloudPod Podcast - where the forecast is always cloudy! Justin, Jonathan, and Ryan are your hosts this week as we discuss all things Google Next! We talk schedule offerings, make our predictions about announcements, and prepare to be generally wrong about everything. Also - do you like stickers? Everyone likes stickers! Be on the lookout for us, and maybe you can have one. Titles we almost went with this week: None! Google Next is the next big thing, so of course it's the title. A big thanks to this week's sponsor: Foghorn Consulting provides top-notch cloud and DevOps engineers to the world's most innovative companies. Initiatives stalled because you have trouble hiring? Foghorn can be burning down your DevOps and Cloud backlogs as soon as next week.
Sloppy practises by Gigabyte reveal one of the problems with UEFI, why Slack refuses to implement end to end encryption, a familiar bug ruins people's uptime, and XFS vs ext4. Plugs Support us on patreon FreeBSD or Linux – A Choice Without OS Wars News/discussion Millions of Gigabyte Motherboards Were Sold With a […]
最大128コア！ AMDが第4世代EPYCに「多コア全振りモデル」と「L3キャッシュ爆増しモデル」を追加。 AMDは6月13日（米国太平洋夏時間）、データセンター向けCPU「第4世代EPYCプロセッサ」のスケーラビリティー重視モデル（開発コード名：Bergamo）と、3D V-Cacheテクノロジーを適用してL3キャッシュを増量したモデル（開発コード名：Genoa-X）の出荷開始を発表した。
Wendell joins to discuss Xeon, EPYC, 13th Gen Releases, Threadripper, and more! [SPON: dieshrink = 3% off Everything, brokensilicon = 25% off Windows: https://biitt.ly/shbSk ] [SPON: Get 10% off Vite Ramen AND a FREE Pack w/ “MOORESLAW”: https://bit.ly/3wKx6v1 ] [SPON: Get 6% OFF Custom PCs & GPUs at https://www.silverknightpcs.com/ w/ brokensilicon] 0:00 Who is Wendell? What is a “Computer Janitor”? 10:27 i9-13900KS, i5-13500, DDR5-7200 Thoughts 16:30 Does Intel have anything to counter Zen 4 X3D & Ryzen non-X? 30:51 Sapphire Rapids - Is it actually competitive with Genoa? 38:13 Why Netflix and other customers don't want SPR's Accelerators… 48:09 Will licensing costs doom SPR? Would Wendell trade cores for accelerators? 58:53 Bergamo, Genoa-X, Siena Launch 2023 – Will SPR be Quadruple Teamed? 1:07:51 Can AMD get to 40% Server Market Share? Can Granite Rapids stop it? 1:16:16 Is AMD not being aggressive enough with Laptop and Threadripper? 1:22:27 Is AMD ROCm ready to take on Nvidia CUDA? 1:35:04 Threadripper vs Intel Fishhawk Falls 1:41:55 AMD's Strategy to take Laptop Market Share 1:52:31 6GHz Sever Chips, MI300, All-Cache Memory, X86 vs ARM 2:04:56 2023-2024 are Critical Years for determining Intel's Future… 2:20:20 Don't Forget – Incredible Performance Uplifts are Happening Yearly! https://www.youtube.com/@Level1Techs Wendell's i9-13900KS Review: https://youtu.be/rBexggqldDs Latest MLID Xeon Leak: https://youtu.be/h20inMLeDnE MLID Leak of Genoa Accelerators: https://youtu.be/6yFn85I5PbY?t=878 https://www.amazon.com/Best-Sellers-Computer-CPU-Processors/zgbs/pc/229189 https://www.tomshardware.com/news/intel-launches-sapphire-rapids-fourth-gen-xeon-cpus-and-ponte-vecchio-max-gpu-series https://www.intel.com/content/www/us/en/products/sku/212287/intel-xeon-platinum-8380-processor-60m-cache-2-30-ghz/specifications.html https://ark.intel.com/content/www/us/en/ark/products/134586/intel-core-i512400-processor-18m-cache-up-to-4-40-ghz.html https://www.intel.com/content/www/us/en/products/sku/230580/intel-core-i513500-processor-24m-cache-up-to-4-80-ghz/specifications.html https://www.screenhacker.com/intel-sapphire-rapids-xeon-4-tile-mcm-annotated/
CXL couldn't be utilized until there was a server platform that supported it, so we're very excited to see AMD launch their next-generation Epyc server platform, code-name Genoa. In this episode of Utilizing CXL, Stephen Foskett and Craig Rodgers talk with Mahesh Wagh of AMD and the CXL Consortium about this important release. The first step to bring CXL to market is to prove it is functional and performs well with a mainstream platform like AMD 4th-generation Epyc. AMD is bringing CXL to market as a tiered memory solution that performs similar to as memory on a remote NUMA socket without any special configuration or software. But AMD also supports other memory technologies, including hierarchical memory with software, security, pooling, and even memory sharing with specialized software. Although Epyc is said to only support CXL 1.1, later spec devices will be backwards-compatible and the platform also supports some CXL 2.0 features including global flush for persistent memory, firmware-first error handling, and device-specific capabilities. Hosts: Stephen Foskett: https://www.twitter.com/SFoskett Craig Rodgers: https://www.twitter.com/CraigRodgersms Guest Host: Mahesh Wagh, Senior Fellow at AMD, Server Systems Architect, Co-chair CXL Consortium Board Technical Task Force https://www.linkedin.com/in/maheshwagh/ Follow Gestalt IT and Utilizing Tech Website: https://www.UtilizingTech.com/ Website: https://www.GestaltIT.com/ Twitter: https://www.twitter.com/GestaltIT LinkedIn: https://www.linkedin.com/company/1789 Tags: #UtilizingCXL #Genoa #Epyc #AMD #CXL
We have made it to episode 700!!! Sure, only half of the crew made it this far, but that's just how it goes. It's a war of attrition, and Jeremy was there from episode one (Josh joined a bit later). And this week we decided to gather, virtually, once again, and talk about the most affordable graphics card ever made!RTX 4080 review, gaming PC building vs consoles, Nvidia earnings, some good security news for once, and more!Timestamps:00:00 Intro01:39 Food with Josh03:08 NVIDIA GeForce RTX 4080 Founders Edition review36:07 You aren't dreaming...it's NVIDIA quarterly earnings coverage!42:34 Podcast sponsor - Masterclass44:03 Did Gamers Nexus solve the 4090 connector mystery?55:47 AMD's EPYC new server processors1:00:53 Josh talks memory controllers1:04:23 AMD brings HYPR-RX to Radeon Software next year1:06:09 Google settles Android tracking lawsuit1:08:02 Lenovo patches UEFI vulnerability1:10:01 Gaming PSA1:13:12 Picks of the Week1:25:45 Outro ★ Support this podcast on Patreon ★
This week we covered: News AMD 4th Gen EPYC 9004 “Genoa Zen 4” CPUs Launched: Up To 96 Cores, 192 Threads, 384 MB L3 Cache & Crushing All Other Server Chips Google will pay $392 million to 40 states in largest-ever US consumer privacy settlement Apple sued for collecting user data despite opt-outs Amazon mass layoffs will reportedly ax 10,000 people this week Twitter lays off 5K contractors in surprise 2nd wave of cuts, more mods lost Musk fires Twitter engineers for correcting, criticizing him on Twitter, Slack Rumours GOOGLE PIXEL FOLD - EXCLUSIVE FIRST LOOK! Wild design, specs, price, and more! As always, we'd love to hear your comments Find us on Twitter @WeeklyTechRant
AMD got the jump on Intel in the battle for fourth-generation server CPUs, announcing the so-called Genoa line at SuperComputing 22. AMD's new CPU line is a massive upgrade, with more cores, accelerator instructions, and CXL. And it comes ahead of Intel's delayed Sapphire Rapids Xeon announcement, which is widely expected early next year. Let's take a closer look at the AMD Genoa CPU line. Time Stamps: 0:00 - Welcome to the Rundown 0:36 - VMware Purchases Ananda Networks 2:38 - NetApp's BlueXP Manages Data Anywhere 4:50 - Kalray Enters the Storage Accelerator Market 7:15 - Hammerspace Hits the Afterburner 9:18 - Alluxio 2.9 Connects Data and Applications 12:34 - AMD Releases Genoa Epyc CPUs 24:46 - The Weeks Ahead 25:49 - Thanks for Watching Follow our hosts on Social Media Tom Hollingsworth: https://www.twitter.com/NetworkingNerd Stephen Foskett: https://www.twitter.com/SFoskett Max Mortillaro: https://www.twitter.com/MaxMortillaro Follow Gestalt IT Website: https://www.GestaltIT.com/ Twitter: https://www.twitter.com/GestaltIT LinkedIn: https://www.linkedin.com/company/1789
來看看和半導體市場有關的消息，近年積極搶攻伺服器市場的AMD超微，10日推出第4代EPYC處理器，核心數量將高達96顆，功耗也大大降低，無論是性能還是效能，都比上一代提升不少。這回同樣由台積電代工生產，採用5奈米製程等。而台積電ADR也有好消息，在費城半導體大漲超過10%的激勵下，同步勁揚8.9%，收在70.84美元。至於半導體微影設備廠艾司摩爾ASML，則在10日宣布，將啓動高達120億歐元的股票回購計劃。 聽眾五星留言+訂閱
最大96コアでライバルを撃墜!? AMDがHPC／データセンター向けCPU「第4世代EPYC」を発売 最大1.9倍のパフォーマンス向上。 AMDは11月10日（米国太平洋時間）、HPC（ハイパフォーマンスコンピューティング）とデータセンターに収容するサーバ向けの新型CPU「EPYC 9004シリーズ」（第4世代EPYC、開発コード名：Genoa）を発表した。製品の出荷は既に開始しており、パートナー企業を通して同CPUを組み込んだシステムなどが順次発売される予定だ。
There are big technological changes happening in the sometimes arcane world of server computing, as Pinnacle ICT's Ricky Pereira and Peter van der Merwe reveal in this episode of the podcast. Pereira, who is Pinnacle's Dell Enterprise product manager, and Van der Merwe, a pre-sales engineer, unpack the new range of Dell EMC PowerEdge servers and the innovations these workhorses are bringing to a data centre near you. Pereira unpacks why these 15th-generation PC servers from Dell are game changers, including: * New CPUs from Intel (latest-generation Xeon) and AMD (latest-generation Epyc) and what they can do; * More cores, resulting in better performance, including in virtualisation - and why this is important; * Improvements to memory and storage; * Faster connectivity with new network controllers; * Liquid cooling as an option, and why this traditionally gamer-focused technology is headed into server farms; * Upgraded Raid controllers; and The automation of configurations for faster deployment and support. The two also discuss the security enhancements that the PowerEdge servers bring to enterprises, and how support and service offerings to the market have been improved. This promoted podcast was paid for by the party concerned
There are big technological changes happening in the sometimes arcane world of server computing, as Pinnacle ICT's Ricky Pereira and Peter van der Merwe reveal in this episode of the podcast. Pereira, who is Pinnacle's Dell Enterprise product manager, and Van der Merwe, a pre-sales engineer, unpack the new range of Dell EMC PowerEdge servers and the innovations these workhorses are bringing to a data centre near you. Pereira unpacks why these 15th-generation PC servers from Dell are game changers, including: * New CPUs from Intel (latest-generation Xeon) and AMD (latest-generation Epyc) and what they can do; * More cores, resulting in better performance, including in virtualisation - and why this is important; * Improvements to memory and storage; * Faster connectivity with new network controllers; * Liquid cooling as an option, and why this traditionally gamer-focused technology is headed into server farms; * Upgraded Raid controllers; and The automation of configurations for faster deployment and support. The two also discuss the security enhancements that the PowerEdge servers bring to enterprises, and how support and service offerings to the market have been improved. This promoted podcast was paid for by the party concerned TechCentral
AMD Fixes Stuttering but is removing OC from the 5800X3D, Apple debuts the M1 Ultra and it's formidable, Threadripper is coming back in, Linux catches a dirty pipe, so many gaming bundles and more topics in the timeline below!Timestamps:00:00 Intro01:50 Burger of the Week05:27 Threadripper is BACK07:36 Apple M1 Ultra13:40 Basemark's new ray tracing benchmark16:13 AMD Ryzen stuttering with firmware TPM?18:25 Podcast Sponsor: New Relic19:37 No OC for Ryzen 7 5800X3D?23:51 3D V-cache EPYC on the way24:28 Linux nobody account bug26:32 NVIDIA and Samsung cyberattack updates32:02 Amazon Echo hack33:06 Intel W680 chipset supports ECC memory on Alder Lake CPUs35:16 Gaming segment: a bunch of bundles40:33 Picks of the Week52:15 Outro★ Support this podcast on Patreon ★
本集節目由 Supermicro 美商美超微贊助播出 Supermicro，美超微電腦，為應用最佳化全面性 IT 解決方案的全球領導者。 Supermicro HIGH PERFORMANCE Building Block產品組合，搭載AMD第3代 EPYC™ 處理器，透過世界領先HPC(超級運算)系統為您建構最佳化IT解決方案 了解更多: https://learn-more.supermicro.com/high-performance-building-blocks?utm_source=Podcast_SoundOn&utm_medium=podcastad&utm_campaign=AMDBuildingBlocks --- 本集主題 - Disney+ 推出廣告版、亞馬遜 Luna 正式上市、加密貨幣援烏克蘭 --- M觀點資訊 --- 科技巨頭解碼: https://bit.ly/3koflbU M觀點 Telegram - https://t.me/miulaviewpoint M觀點 IG - https://www.instagram.com/miulaviewpoint/ M觀點Podcast - https://bit.ly/34fV7so M報: https://bit.ly/345gBbA M觀點YouTube頻道訂閱 https://bit.ly/2nxHnp9 M觀點粉絲團 https://www.facebook.com/miulaperspective/ 任何合作邀約請洽 miula@outlook.com
On The Cloud Pod this week, Jonathan's got his detective hat on. Plus Akamai steps up to CloudFare with Linode acquisition, AWS' CloudFormation Hooks lift us up, and EPYC instances are now available. A big thanks to this week's sponsors: Foghorn Consulting, which provides full-stack cloud solutions with a focus on strategy, planning and execution for enterprises seeking to take advantage of the transformative capabilities of AWS, Google Cloud and Azure. This week's highlights
Twitter launches Blue subscription service in U.S., offers ad-free access to 300 news sites. U.S. states file updated antitrust complaint against Alphabet's Google. Walmart is using fully driverless trucks to ramp up its online grocery business. Amazon's Spinmasters: Behind the Internet Giant's Battle With the Press. Are you worried about Amazon? A record 4.4 million people quit in September as Great Resignation shows no signs of stopping. Job openings and labor turnover trends for States in 2020. Why America's "Shipping Crisis" Will Not End. Amazon Relay. Bezos-Backed Real Estate App to Accept Bitcoin as Payment. Golem Network. The Genesis Machine. The EVELO Compass adult electric tricycle is a comfortable, powerful and stable electric trike with a large cargo basket. Wing Bikes are the world's best electric bicycles. Hoax Email Blast Abused Poor Coding in FBI Website. Europol: Seven REvil/GandCrab ransomware affiliates were arrested in 2021. Judge denies stay in Epic v. Apple. Everyone has questions about Leo's phones. GitHub Copilot: Your AI pair programmer. Apple to Pay $30 Million Over Store Workers' Security Checks. Facebook says it can't keep pace with its own Oversight Board. Elon Musk Sells About $5 Billion in Tesla Stock. AMD reveals an Epyc security 50 flaws, Intel has 25. Samsung Galaxy Buds Pro are infectious in all the wrong ways. Rivian's Mega IPO is a Good Test of the Meme Stock Craze. AMC Theatres debuts online Bitcoin payments after months of teasing. Biden's "historic" $65 billion broadband plan approved by Congress. Host: Leo Laporte Guests: Paris Martineau and Amy Webb Download or subscribe to this show at https://twit.tv/shows/this-week-in-tech Get episodes ad-free with Club TWiT at https://twit.tv/clubtwit Sponsors: itpro.tv/twit promo code TWIT30 blockfi.com/TWIT mintmobile.com/twit wwt.com/twit
Twitter launches Blue subscription service in U.S., offers ad-free access to 300 news sites. U.S. states file updated antitrust complaint against Alphabet's Google. Walmart is using fully driverless trucks to ramp up its online grocery business. Amazon's Spinmasters: Behind the Internet Giant's Battle With the Press. Are you worried about Amazon? A record 4.4 million people quit in September as Great Resignation shows no signs of stopping. Job openings and labor turnover trends for States in 2020. Why America's "Shipping Crisis" Will Not End. Amazon Relay. Bezos-Backed Real Estate App to Accept Bitcoin as Payment. Golem Network. The Genesis Machine. The EVELO Compass adult electric tricycle is a comfortable, powerful and stable electric trike with a large cargo basket. Wing Bikes are the world's best electric bicycles. Hoax Email Blast Abused Poor Coding in FBI Website. Europol: Seven REvil/GandCrab ransomware affiliates were arrested in 2021. Judge denies stay in Epic v. Apple. Everyone has questions about Leo's phones. GitHub Copilot: Your AI pair programmer. Apple to Pay $30 Million Over Store Workers' Security Checks. Facebook says it can't keep pace with its own Oversight Board. Elon Musk Sells About $5 Billion in Tesla Stock. AMD reveals an Epyc security 50 flaws, Intel has 25. Samsung Galaxy Buds Pro are infectious in all the wrong ways. Rivian's Mega IPO is a Good Test of the Meme Stock Craze. AMC Theatres debuts online Bitcoin payments after months of teasing. Biden's "historic" $65 billion broadband plan approved by Congress. Host: Leo Laporte Guests: Paris Martineau and Amy Webb Download or subscribe to this show at https://twit.tv/shows/this-week-in-tech Get episodes ad-free with Club TWiT at https://twit.tv/clubtwit Sponsors: itpro.tv/twit promo code TWIT30 blockfi.com/TWIT mintmobile.com/twit wwt.com/twit
Twitter launches Blue subscription service in U.S., offers ad-free access to 300 news sites. U.S. states file updated antitrust complaint against Alphabet's Google. Walmart is using fully driverless trucks to ramp up its online grocery business. Amazon's Spinmasters: Behind the Internet Giant's Battle With the Press. Are you worried about Amazon? A record 4.4 million people quit in September as Great Resignation shows no signs of stopping. Job openings and labor turnover trends for States in 2020. Why America's "Shipping Crisis" Will Not End. Amazon Relay. Bezos-Backed Real Estate App to Accept Bitcoin as Payment. Golem Network. The Genesis Machine. The EVELO Compass adult electric tricycle is a comfortable, powerful and stable electric trike with a large cargo basket. Wing Bikes are the world's best electric bicycles. Hoax Email Blast Abused Poor Coding in FBI Website. Europol: Seven REvil/GandCrab ransomware affiliates were arrested in 2021. Judge denies stay in Epic v. Apple. Everyone has questions about Leo's phones. GitHub Copilot: Your AI pair programmer. Apple to Pay $30 Million Over Store Workers' Security Checks. Facebook says it can't keep pace with its own Oversight Board. Elon Musk Sells About $5 Billion in Tesla Stock. AMD reveals an Epyc security 50 flaws, Intel has 25. Samsung Galaxy Buds Pro are infectious in all the wrong ways. Rivian's Mega IPO is a Good Test of the Meme Stock Craze. AMC Theatres debuts online Bitcoin payments after months of teasing. Biden's "historic" $65 billion broadband plan approved by Congress. Host: Leo Laporte Guests: Paris Martineau and Amy Webb Download or subscribe to this show at https://twit.tv/shows/this-week-in-tech Get episodes ad-free with Club TWiT at https://twit.tv/clubtwit Sponsors: itpro.tv/twit promo code TWIT30 blockfi.com/TWIT mintmobile.com/twit wwt.com/twit
Twitter launches Blue subscription service in U.S., offers ad-free access to 300 news sites. U.S. states file updated antitrust complaint against Alphabet's Google. Walmart is using fully driverless trucks to ramp up its online grocery business. Amazon's Spinmasters: Behind the Internet Giant's Battle With the Press. Are you worried about Amazon? A record 4.4 million people quit in September as Great Resignation shows no signs of stopping. Job openings and labor turnover trends for States in 2020. Why America's "Shipping Crisis" Will Not End. Amazon Relay. Bezos-Backed Real Estate App to Accept Bitcoin as Payment. Golem Network. The Genesis Machine. The EVELO Compass adult electric tricycle is a comfortable, powerful and stable electric trike with a large cargo basket. Wing Bikes are the world's best electric bicycles. Hoax Email Blast Abused Poor Coding in FBI Website. Europol: Seven REvil/GandCrab ransomware affiliates were arrested in 2021. Judge denies stay in Epic v. Apple. Everyone has questions about Leo's phones. GitHub Copilot: Your AI pair programmer. Apple to Pay $30 Million Over Store Workers' Security Checks. Facebook says it can't keep pace with its own Oversight Board. Elon Musk Sells About $5 Billion in Tesla Stock. AMD reveals an Epyc security 50 flaws, Intel has 25. Samsung Galaxy Buds Pro are infectious in all the wrong ways. Rivian's Mega IPO is a Good Test of the Meme Stock Craze. AMC Theatres debuts online Bitcoin payments after months of teasing. Biden's "historic" $65 billion broadband plan approved by Congress. Host: Leo Laporte Guests: Paris Martineau and Amy Webb Download or subscribe to this show at https://twit.tv/shows/this-week-in-tech Get episodes ad-free with Club TWiT at https://twit.tv/clubtwit Sponsors: itpro.tv/twit promo code TWIT30 blockfi.com/TWIT mintmobile.com/twit wwt.com/twit
Twitter launches Blue subscription service in U.S., offers ad-free access to 300 news sites. U.S. states file updated antitrust complaint against Alphabet's Google. Walmart is using fully driverless trucks to ramp up its online grocery business. Amazon's Spinmasters: Behind the Internet Giant's Battle With the Press. Are you worried about Amazon? A record 4.4 million people quit in September as Great Resignation shows no signs of stopping. Job openings and labor turnover trends for States in 2020. Why America's "Shipping Crisis" Will Not End. Amazon Relay. Bezos-Backed Real Estate App to Accept Bitcoin as Payment. Golem Network. The Genesis Machine. The EVELO Compass adult electric tricycle is a comfortable, powerful and stable electric trike with a large cargo basket. Wing Bikes are the world's best electric bicycles. Hoax Email Blast Abused Poor Coding in FBI Website. Europol: Seven REvil/GandCrab ransomware affiliates were arrested in 2021. Judge denies stay in Epic v. Apple. Everyone has questions about Leo's phones. GitHub Copilot: Your AI pair programmer. Apple to Pay $30 Million Over Store Workers' Security Checks. Facebook says it can't keep pace with its own Oversight Board. Elon Musk Sells About $5 Billion in Tesla Stock. AMD reveals an Epyc security 50 flaws, Intel has 25. Samsung Galaxy Buds Pro are infectious in all the wrong ways. Rivian's Mega IPO is a Good Test of the Meme Stock Craze. AMC Theatres debuts online Bitcoin payments after months of teasing. Biden's "historic" $65 billion broadband plan approved by Congress. Host: Leo Laporte Guests: Paris Martineau and Amy Webb Download or subscribe to this show at https://twit.tv/shows/this-week-in-tech Get episodes ad-free with Club TWiT at https://twit.tv/clubtwit Sponsors: itpro.tv/twit promo code TWIT30 blockfi.com/TWIT mintmobile.com/twit wwt.com/twit
Obviously the Steam Deck falls victim to the supply chain problems, AMD EPYC announcements and how big they're getting in the datacenter, Cisco switch vulnerabilities, review of the Lexar DDR4 3600, why Intel wanted Centaur, and beQuiet! intros some RGB fans. Plus all the "more". You're here for that.Another week is in the books, much more in a pure stream of consciousness that was originally streamed live, and has now been edited for human consumption. Full list of subjects in the time stamps list below.Timestamps0:00 Intro01:19 Burger of the Week03:30 Steam Deck Delayed06:15 AMD's EPYC Data Center Event26:41 Podcast Sponsor: Comet Backup27:58 Why Intel Wanted Centaur33:04 Alder Lake Has AVX-512 Support?34:42 Some Cisco Switch Vulnerabilities37:33 Podcast Sponsor: VPLS 39:01 be quiet! Intros Company's First ARGB Fans41:38 The Stylish Raijintek PAN SLIM Case46:36 Lexar Hades RGB DDR4 3600 RAM51:19 Picks of the Week★ Support this podcast on Patreon ★
0:00 help 0:06 EPYC Bergamo 128-core CPU 1:03 Apple Silicon 3nm 2:04 Pixel 3 charging sucks 2:54 Vessi Footwear 3:28 QUICK BITS 3:35 Nintendo Switch 2 coming...eventually 4:18 Win7, 8.1 OneDrive support ending 4:46 Wind and solar energy r good? 5:18 Walmart driverless deliveries 5:43 McDonald's uses IBM AI News Sources: https://lmg.gg/HDtD5
Microsoft and Ubuntu's relationship is under a new spotlight this week. Plus our rundown of the feature-packed 5.11 release, a Fuchsia surprise, exciting hardware news, and more.
Patrick Norton is joined by Jim Tanous, Managing Editor at PC Perspective, to chat about AMD's EPYC 7742 server processor, the encoding company Beamr, Corsair's VENGEANCE LPX DDR4 Memory, Intel's Ice Lake graphics leap, the Core i9-9900KS TDP leak, Huawei's new Mate 30 Pro phone, Facebook's 2nd-gen Portal products, and more! Host: Patrick Norton Guest: Jim Tanous Download or subscribe to this show at https://twit.tv/shows/this-week-in-computer-hardware. Send your computer hardware questions to twich@twit.tv. Sponsor: plex.tv/twit code TWIT10
In this episode, we take a look at the reimplementation of NetBSD using a Microkernel, check out what makes DHCP faster, and see what high-process count support for DragonflyBSD has to offer, and we answer the questions you've always wanted to ask us. This episode was brought to you by Headlines A Reimplementation Of Netbsd Using a Microkernel (http://theembeddedboard.review/a-reimplementation-of-netbsd-using-a-microkernel-part-1-of-2/) Minix author Andy Tanenbaum writes in Part 1 of a-reimplementation-of-netbsd-using-a-microkernel (http://theembeddedboard.review/a-reimplementation-of-netbsd-using-a-microkernel-part-1-of-2/) Based on the MINIX 3 microkernel, we have constructed a system that to the user looks a great deal like NetBSD. It uses pkgsrc, NetBSD headers and libraries, and passes over 80% of the KYUA tests). However, inside, the system is completely different. At the bottom is a small (about 13,000 lines of code) microkernel that handles interrupts, message passing, low-level scheduling, and hardware related details. Nearly all of the actual operating system, including memory management, the file system(s), paging, and all the device drivers run as user-mode processes protected by the MMU. As a consequence, failures or security issues in one component cannot spread to other ones. In some cases a failed component can be replaced automatically and on the fly, while the system is running, and without user processes noticing it. The talk will discuss the history, goals, technology, and status of the project. Research at the Vrije Universiteit has resulted in a reimplementation of NetBSD using a microkernel instead of the traditional monolithic kernel. To the user, the system looks a great deal like NetBSD (it passes over 80% of the KYUA tests). However, inside, the system is completely different. At the bottom is a small (about 13,000 lines of code) microkernel that handles interrupts, message passing, low-level scheduling, and hardware related details. Nearly all of the actual operating system, including memory management, the file system(s), paging, and all the device drivers run as user-mode processes protected by the MMU. As a consequence, failures or security issues in one component cannot spread to other ones. In some cases a failed component can be replaced automatically and on the fly, while the system is running. The latest work has been adding live update, making it possible to upgrade to a new version of the operating system WITHOUT a reboot and without running processes even noticing. No other operating system can do this. The system is built on MINIX 3, a derivative of the original MINIX system, which was intended for education. However, after the original author, Andrew Tanenbaum, received a 2 million euro grant from the Royal Netherlands Academy of Arts and Sciences and a 2.5 million euro grant from the European Research Council, the focus changed to building a highly reliable, secure, fault tolerant operating system, with an emphasis on embedded systems. The code is open source and can be downloaded from www.minix3.org. It runs on the x86 and ARM Cortex V8 (e.g., BeagleBones). Since 2007, the Website has been visited over 3 million times and the bootable image file has been downloaded over 600,000 times. The talk will discuss the history, goals, technology, and status of the project. Part 2 (http://theembeddedboard.review/a-reimplementation-of-netbsd-using-a-microkernel-part-2-of-2/) is also available. *** Rapid DHCP: Or, how do Macs get on the network so fast? (https://cafbit.com/post/rapid_dhcp_or_how_do/) One of life's minor annoyances is having to wait on my devices to connect to the network after I wake them from sleep. All too often, I'll open the lid on my EeePC netbook, enter a web address, and get the dreaded "This webpage is not available" message because the machine is still working on connecting to my Wi-Fi network. On some occasions, I have to twiddle my thumbs for as long as 10-15 seconds before the network is ready to be used. The frustrating thing is that I know it doesn't have to be this way. I know this because I have a Mac. When I open the lid of my MacBook Pro, it connects to the network nearly instantaneously. In fact, no matter how fast I am, the network comes up before I can even try to load a web page. My curiosity got the better of me, and I set out to investigate how Macs are able to connect to the network so quickly, and how the network connect time in other operating systems could be improved. I figure there are three main categories of time-consuming activities that occur during network initialization: Link establishment. This is the activity of establishing communication with the network's link layer. In the case of Wi-Fi, the radio must be powered on, the access point detected, and the optional encryption layer (e.g. WPA) established. After link establishment, the device is able to send and receive Ethernet frames on the network. Dynamic Host Configuration Protocol (DHCP). Through DHCP handshaking, the device negotiates an IP address for its use on the local IP network. A DHCP server is responsible for managing the IP addresses available for use on the network. Miscellaneous overhead. The operating system may perform any number of mundane tasks during the process of network initialization, including running scripts, looking up preconfigured network settings in a local database, launching programs, etc. My investigation thus far is primarily concerned with the DHCP phase, although the other two categories would be interesting to study in the future. I set up a packet capture environment with a spare wireless access point, and observed the network activity of a number of devices as they initialized their network connection. For a worst-case scenario, let's look at the network activity captured while an Android tablet is connecting: This tablet, presumably in the interest of "optimization", is initially skipping the DHCP discovery phase and immediately requesting its previous IP address. The only problem is this is a different network, so the DHCP server ignores these requests. After about 4.5 seconds, the tablet stubbornly tries again to request its old IP address. After another 4.5 seconds, it resigns itself to starting from scratch, and performs the DHCP discovery needed to obtain an IP address on the new network. In all fairness, this delay wouldn't be so bad if the device was connecting to the same network as it was previously using. However, notice that the tablet waits a full 1.13 seconds after link establishment to even think about starting the DHCP process. Engineering snappiness usually means finding lots of small opportunities to save a few milliseconds here and there, and someone definitely dropped the ball here. In contrast, let's look at the packet dump from the machine with the lightning-fast network initialization, and see if we can uncover the magic that is happening under the hood: The key to understanding the magic is the first three unicast ARP requests. It looks like Mac OS remembers certain information about not only the last connected network, but the last several networks. In particular, it must at least persist the following tuple for each of these networks: > 1. The Ethernet address of the DHCP server > 2. The IP address of the DHCP server > 3. Its own IP address, as assigned by the DHCP server During network initialization, the Mac transmits carefully crafted unicast ARP requests with this stored information. For each network in its memory, it attempts to send a request to the specific Ethernet address of the DHCP server for that network, in which it asks about the server's IP address, and requests that the server reply to the IP address which the Mac was formerly using on that network. Unless network hosts have been radically shuffled around, at most only one of these ARP requests will result in a response—the request corresponding to the current network, if the current network happens to be one of the remembered networks. This network recognition technique allows the Mac to very rapidly discover if it is connected to a known network. If the network is recognized (and presumably if the Mac knows that the DHCP lease is still active), it immediately and presumptuously configures its IP interface with the address it knows is good for this network. (Well, it does perform a self-ARP for good measure, but doesn't seem to wait more than 13ms for a response.) The DHCP handshaking process begins in the background by sending a DHCP request for its assumed IP address, but the network interface is available for use during the handshaking process. If the network was not recognized, I assume the Mac would know to begin the DHCP discovery phase, instead of sending blind requests for a former IP address as the Galaxy Tab does. The Mac's rapid network initialization can be credited to more than just the network recognition scheme. Judging by the use of ARP (which can be problematic to deal with in user-space) and the unusually regular transmission intervals (a reliable 1.0ms delay between each packet sent), I'm guessing that the Mac's DHCP client system is entirely implemented as tight kernel-mode code. The Mac began the IP interface initialization process a mere 10ms after link establishment, which is far faster than any other device I tested. Android devices such as the Galaxy Tab rely on the user-mode dhclient system (part of the dhcpcd package) dhcpcd program, which no doubt brings a lot of additional overhead such as loading the program, context switching, and perhaps even running scripts. The next step for some daring kernel hacker is to implement a similarly aggressive DHCP client system in the Linux kernel, so that I can enjoy fast sign-on speeds on my Android tablet, Android phone, and Ubuntu netbook. There already exists a minimal DHCP client implementation in the Linux kernel, but it lacks certain features such as configuring the DNS nameservers. Perhaps it wouldn't be too much work to extend this code to support network recognition and interface with a user-mode daemon to handle such auxillary configuration information received via DHCP. If I ever get a few spare cycles, maybe I'll even take a stab at it. You can also find other ways of optimizing the dhclient program and how it works in the dhclient tutorial on Calomel.org (https://calomel.org/dhclient.html). *** BSDCam Trip Report (https://www.freebsdfoundation.org/blog/bsdcam-2017-trip-report-michael-lucas/) Over the decades, FreeBSD development and coordination has shifted from being purely on-line to involving more and more in-person coordination and cooperation. The FreeBSD Foundation sponsors a devsummit right before BSDCan, EuroBSDCon, and AsiaBSDCon, so that developers traveling to the con can leverage their airfare and hammer out some problems. Yes, the Internet is great for coordination, but nothing beats a group of developers spending ten minutes together to sketch on a whiteboard and figuring out exactly how to make something bulletproof. In addition to the coordination efforts, though, conference devsummits are hierarchical. There's a rigid schedule, with topics decided in advance. Someone leads the session. Sessions can be highly informative, passionate arguments, or anything in between. BSDCam is… a little different. It's an invaluable part of the FreeBSD ecosystem. However, it's something that I wouldn't normally attend. But right now, is not normal. I'm writing a new edition of Absolute FreeBSD. To my astonishment, people have come to rely on this book when planning their deployments and operations. While I find this satisfying, it also increases the pressure on me to get things correct. When I wrote my first FreeBSD book back in 2000, a dozen mailing lists provided authoritative information on FreeBSD development. One person could read every one of those lists. Today, that's not possible—and the mailing lists are only one narrow aspect of the FreeBSD social system. Don't get me wrong—it's pretty easy to find out what people are doing and how the system works. But it's not that easy to find out what people will be doing and how the system will work. If this book is going to be future-proof, I needed to leave my cozy nest and venture into the wilds of Cambridge, England. Sadly, the BSDCam chair agreed with my logic, so I boarded an aluminum deathtrap—sorry, a “commercial airliner”—and found myself hurtled from Detroit to Heathrow. And one Wednesday morning, I made it to the William Gates building of Cambridge University, consciousness nailed to my body by a thankfully infinite stream of proper British tea. BSDCam attendance is invitation only, and the facilities can only handle fifty folks or so. You need to be actively working on FreeBSD to wrangle an invite. Developers attend from all over the world. Yet, there's no agenda. Robert Watson is the chair, but he doesn't decide on the conference topics. He goes around the room and asks everyone to introduce themselves, say what they're working on, and declare what they want to discuss during the conference. The topics of interest are tallied. The most popular topics get assigned time slots and one of the two big rooms. Folks interested in less popular topics are invited to claim one of the small breakout rooms. Then the real fun begins. I started by eavesdropping in the virtualization workshop. For two hours, people discussed FreeBSD's virtualization needs, strengths, and weaknesses. What needs help? What should this interface look like? What compatibility is important, and what isn't? By the end of the session, the couple dozen people had developed a reasonable consensus and, most importantly, some folks had added items to their to-do lists. Repeat for a dozen more topics. I got a good grip on what's really happening with security mitigation techniques, FreeBSD's cloud support, TCP/IP improvements, advances in teaching FreeBSD, and more. A BSDCan devsummit presentation on packaging the base system is informative, but eavesdropping on two dozen highly educated engineers arguing about how to nail down the final tidbits needed to make that a real thing is far more educational. To my surprise, I was able to provide useful feedback for some sessions. I speak at a lot of events outside of the FreeBSD world, and was able to share much of what I hear at Linux conferences. A tool that works well for an experienced developer doesn't necessarily work well for everyone. Every year, I leave BSDCan tired. I left BSDCam entirely exhausted. These intense, focused discussions stretched my brain. But, I have a really good idea where key parts of FreeBSD development are actually headed. This should help future-proof the new Absolute FreeBSD, as much as any computer book can be future-proof. Plus, BSDCam throws the most glorious conference dinner I've ever seen. I want to thank Robert Watson for his kind invitation, and the FreeBSD Foundation for helping defray the cost of this trip Interview - The BSDNow Crew As a kid, what did you dream of to become as an adult? JT: An Astronaut BR: I wanted to be a private detective, because of all the crime novels that I read back then. I didn't get far with it. However, I think the structured analysis skills (who did what, when, and such) help me in debugging and sysadmin work. AJ: Didn't think about it much How do you manage to stay organized day to day with so much things you're actively doing each day? (Day job, wife/girlfriend, conferences, hobbies, friends, etc.) JT: Who said I was organized? BR: A lot of stuff in my calendar as reminders, open browser tabs as “to read later” list. A few things like task switching when getting stuck helps. Also, focus on a single goal for the day, even though there will be distractions. Slowly, but steadily chip away at the things you're working on. Rather than to procrastinate and put things back to review later, get started early with easy things for a big task and then tackle the hard part. Often, things look totally chaotic and unmanageable, until you start working on them. AJ: I barely manage. Lots of Google Calendar reminders, and the entire wall of my office is covered in whiteboard sheet todo lists. I use pinboard.in to deal with finding and organizing bookmarks. Write things down, don't trust your memory. What hobbies outside of IT do you have? JT: I love photography, but I do that Professional part time, so I'm not sure if that counts as a hobby anymore. I guess it'd have to be working in the garage on my cars. BR: I do Tai Chi to relax once a week in a group, but can also do it alone, pretty much everywhere. Way too much Youtube watching and browsing the web. I did play some games before studying at the university and I'm still proud that I could control it to the bare minimum not to impact my studies. A few “lapses” from time to time, revisiting the old classics since the newer stuff won't run on my machines anyway. Holiday time is pretty much spent for BSD conferences and events, this is where I can relax and talk with like-minded people from around the world, which is fascinating. Plus, it gets me to various places and countries I never would have dared to visit on my own. AJ: I play a few video games, and I like to ski, although I don't go very often as most of my vacation time is spent hanging out with my BSD friends at various conferences How do you relax? JT: What is this word ‘relax' and what does it mean? BR: My Tai Chi plays a big part in it I guess. I really calms you and the constant stream of thoughts for a while. It also gives you better clarity of what's important in life. Watching movies, sleeping long. AJ: Usually watching TV or Movies. Although I have taken to doing most of my TV watching on my exercise bike now, but it is still mentally relaxing If FreeBSD didn't exist, which BSD flavour would you use? Why? JT: I use TrueOS, but if FreeBSD didn't exist, that project might not either… so… My other choice would be HardenedBSD, but since it's also based on FreeBSD I'm in the same dillema. BR: I once installed NetBSD to see what It can do. If FreeBSD wouldn't exist, I would probably try my luck with it. OpenBSD is also appealing, but I've never installed it. AJ: When I started using FreeBSD in 2000, the only other BSD I had heard of at the time was OpenBSD. If FreeBSD wasn't around, I don't think the world would look like it does, so it is hard to speculate. If any of the BSD's weren't around and you had to use Linux, which camp would belong to? (Redhat, SUSE, Debian, Ubuntu, Gentoo?) JT: I learned Linux in the mid 90s using Slackware, which I used consistently up until the mid 2000s, when I joined the PuppyLinux community and eventually became a developer (FYI, Puppy was/is/can be based on Slackware -- its complicated). So I'd go back to using either Slackware or PuppyLinux. BR: I tried various Linux distributions until I landed at Debian. I used is pretty extensively as my desktop OS at home, building custom kernels and packages to install them until I discovered FreeBSD. I ran both side by side for a few months for learning until one day I figured out that I had not booted Debian in a while, so I switched completely. AJ: The first Linux I played with was Slackware, and it is the most BSD like, but the bits of Linux I learned in school were Redhat and so I can somewhat wrap my head around it, although now that they are changing everything to systemd, all of that old knowledge is more harmful than useful. Are you still finding yourself in need to use Windows/Mac OS? Why? JT: I work part time as a professional Photographer, so I do use Windows for my photography work. While I can do everything I need to do in Linux, it comes down to being pragmatic about my time. What takes me several hours to accomplish in Linux I can accomplish in 20 minutes on Windows. BR: I was a long time Windows-only user before my Unix days. But back when Vista was about to come out and I needed a new laptop, my choice was basically learning to cope with Vistas awful features or learn MacOS X. I did the latter, it increased my productivity since it's really a good Unix desktop experience (at least, back then). I only have to use Windows at work from time to time as I manage our Windows Terminal server, which keeps the exposure low enough and I only connect to it to use a certain app not available for the Mac or the BSDs. AJ: I still use Windows to play games, for a lot of video conferencing, and to produce BSD Now. Some of it could be done on BSD but not as easily. I have promised myself that I will switch to 100% BSD rather than upgrade to Windows 10, so we'll see how that goes. Please describe your home networking setup. Router type, router OS, router hardware, network segmentation, wifi apparatus(es), other devices connected, and anything else that might be interesting about your home network. BR: Very simple and boring: Apple Airport Express base station and an AVM FritzBox for DNS, DHCP, and the link to my provider. A long network cable to my desktop machine. That I use less and less often. I just bought an RPI 3 for some home use in the future to replace it. Mostly my brother's and my Macbook Pro's are connected, our phones and the iPad of my mother. AJ: I have a E3-1220 v3 (dual 3.1ghz + HT) with 8 GB of ram, and 4x Intel gigabit server NICs as my router, and it runs vanilla FreeBSD (usually some snapshot of -current). I have 4 different VLANs, Home, Office, DMZ, and Guest WiFi. WiFi is served via a tiny USB powered device I bought in Tokyo years ago, it serves 3 different SSIDs, one for each VLAN except the DMZ. There are ethernet jacks in every room wired for 10 gigabit, although the only machines with 10 gigabit are my main workstation, file server, and some machines in the server rack. There are 3 switches, one for the house (in the laundry room), one for the rack, and one for 10gig stuff. There is a rack in the basement spare bedroom, it has 7 servers in it, mostly storage for live replicas of customer data for my company. How do guys manage to get your work done on FreeBSD desktops? What do you do when you need to a Linux or Windows app that isn't ported, or working? I've made several attempts to switch to FreeBSD, but each attempt failed because of tools not being available (e.g. Zoom, Dropbox, TeamViewer, Crashplan) or broken (e.g. VirtualBox). BR: I use VIrtualBox for everything that is not natively available or Windows-only. Unfortunately, that means no modern games. I mostly do work in the shell when I'm on FreeBSD and when it has to be a graphical application, then I use Fluxbox as the DE. I want to get work done, not look at fancy eye-candy that get's boring after a while. Deactivated the same stuff on my mac due to the same reason. I look for alternative software online, but my needs are relatively easy to satisfy as I'm not doing video editing/rendering and such. AJ: I generally find that I don't need these apps. I use Firefox, Thunderbird, OpenSSH, Quassel, KomodoEdit, and a few other apps, so my needs are not very demanding. It is annoying when packages are broken, but I usually work around this with boot environments, and being able to just roll back to a version that worked for a few days until the problem is solved. I do still have access to a windows machine for the odd time I need specific VPN software or access to Dell/HP etc out-of-band management tools. Which desktop environments are your favorite, and why? For example, I like i3, Xfce, and I'm drawn to Lumina's ethos, but so far always seem to end up back on Xfc because of its ease of use, flexibility, and dashing good looks. JT: As a Lumina Desktop developer, I think my preference is obvious. ;) I am also a long timeOpenBox user, so I have a soft place in my heart for that as well. BR: I use Fluxbox when I need to work with a lot of windows or an application demands X11. KDE and others are too memory heavy for me and I rarely use even 20% of the features they provide. AJ: I was a long time KDE user, but I have adopted Lumina. I find it fast, and that it gets out of my way and lets me do what I want. It had some annoyances early on, but I've nagged the developers into making it work for me. Which command-line shells do you prefer, why, and how (if at all) have you customised the environment or prompt? BR: I use zsh, but without all the fancy stuff you can find online. It might make you more productive, yes. But again, I try to keep things simple. I'm slowly learning tmux and want to work more in it in the future. I sometimes look at other BSD people's laptops and am amazed at what they do with window-management in tmux. My prompt looks like this: bcr@Voyager:~> 20:20 17-08-17 Put this in your .zshrc to get the same result: PROMPT='%n@%m:%~>' RPROMPT='%T %D' AJ: I started using tcsh early on, because it was the shell on the first box I had access to, and because one of the first things I read in “BSD Hacks” was how to enable ‘typo correction”, which made my life a lot better especially on dial up in the early days. My shell prompt looks like this: allan@CA-TOR1-02:/usr/home/allan% What is one thing (or more) missing in FreeBSD you would import from another project or community? Could be tech, process, etc. JT: AUFS from Linux BR: Nohup from Illumos where you can detach an already running process and put it in the background. I often forget that and I'm not in tmux when that happens, so I can see myself use that feature a lot. AJ: Zones (more complete Jails) from IllumOS how do you manage your time to learn about and work on FreeBSD? Does your work/employment enable what you do, or are your contributions mainly done in private time? JT: These days I'm mostly learning things I need for work, so it just falls into something I'm doing while working on work projects. BR: We have a lot of time during the semester holidays to learn on our own, it's part of the idea of being in a university to keep yourself updated, at least for me. Especially in the fast moving world of IT. I also read a lot in my free time. My interests can shift sometimes, but then I devour everything I can find on the topic. Can be a bit excessive, but has gotten me where I am now and I still need a lot to learn (and want to). Since I work with FreeBSD at work (my owndoing), I can try out many things there. AJ: My work means a spend a lot of time working with FreeBSD, but not that much time working ON it. My contributions are mostly done outside of work, but as I own the company I do get more flexibility to take time off for conferences and other FreeBSD related stuff. we know we can bribe Michael W Lucas with gelato (good gelato that is), but what can we use to bribe you guys? Like when I want to have Allan to work on fixing a bug which prevents me from running ZFS on this fancy rock64 board? BR: Desserts of various kinds. AJ: I am probably not the right person to look at your rock64 board. Most people in the project have taken to bribing me with chocolate. In general, my todo list is so long, the best way is a trade, you take this task and I'll take that task. Is your daily mobile device iOS, Android, Windows Mobile, or other? Why? JT: These days I'm using Android on my Blackberry Priv, but until recently I was still a heavy user of Sailfish OS. I would use SailfishOS everyday, if I could find a phone with a keyboard that I could run it on. BR: iOS on the iPhone 7 currently. Never used an Android phone, saw it on other people's devices and what they can do with it (much more). But the infrequent security updates (if any at all) keep me away from it. AJ: I have a Google Nexus 6 (Android 7.1). I wanted the ‘pure' Android experience, and I had been happy with my previous Nexus S. I don't run a custom OS/ROM or anything because I use the phone to verify that video streams work on an ‘average users device'. I am displeased that support for my device will end soon. I am not sure what device I will get next, but it definitely won't be an iPhone. News Roundup Beta Update - Request for (more) Testing (http://undeadly.org/cgi?action=article&sid=20170808065718&mode=flat&count=30) https://beta.undeadly.org/ has received an update. The most significant changes include: The site has been given a less antiquated "look". (As the topic icons have been eliminated, we are no longer seeking help with those graphics.) The site now uses a moderate amount of semantic HTML5. Several bugs in the HTML fragment validator (used for submissions and comments) have been fixed. To avoid generating invalid HTML, submission content which fails validation is no longer displayed in submission/comment previews. Plain text submissions are converted to HTML in a more useful fashion. (Instead of just converting each EOL to , the converter now generates proper paragraphs and interprets two or more consecutive EOLs as indicating a paragraph break.) The redevelopment remains a work-in-progress. Many thanks to those who have contributed! As before, constructive feedback would be appreciated. Of particular interest are reports of bugs in behaviour (for example, in the HTML validator or in authentication) that would preclude the adoption of the current code for the main site. High-process-count support added to master (http://lists.dragonflybsd.org/pipermail/users/2017-August/313552.html) We've fixed a number of bottlenecks that can develop when the number of user processes runs into the tens of thousands or higher. One thing led to another and I said to myself, "gee, we have a 6-digit PID, might as well make it work to a million!". With the commits made today, master can support at least 900,000 processes with just a kern.maxproc setting in /boot/loader.conf, assuming the machine has the memory to handle it. And, in fact, as today's machines start to ratchet up there in both memory capacity and core count, with fast storage (NVMe) and fast networking (10GigE and higher), even in consumer boxes, this is actually something that one might want to do. With AMD's threadripper and EPYC chips now out, the IntelAMD cpu wars are back on! Boasting up to 32 cores (64 threads) per socket and two sockets on EPYC, terabytes of ram, and motherboards with dual 10GigE built-in, the reality is that these numbers are already achievable in a useful manner. In anycase, I've tested these changes on a dual-socket xeon. I can in-fact start 900,000 processes. They don't get a whole lot of cpu and running 'ps' would be painful, but it works and the system is still responsive from the shell with all of that going on. xeon126# uptime 1:42PM up 9 mins, 3 users, load averages: 890407.00, 549381.40, 254199.55 In fact, judging from the memory use, these minimal test processes only eat around 60KB each. 900,000 of them ate only 55GB on a 128GB machine. So even a million processes is not out of the question, depending on the cpu requirements for those processes. Today's modern machines can be stuffed with enormous amounts of memory. Of course, our PIDs are currently limited to 6 digits, so a million is kinda the upper limit in terms of discrete user processes (verses pthreads which are less restricted). I'd rather not go to 7 digits (yet). CFT: Driver for generic MS Windows 7/8/10 - compatible USB HID multi-touch touchscreens (https://lists.freebsd.org/pipermail/freebsd-current/2017-August/066783.html) Following patch [1] adds support for generic MS Windows 7/8/10 - compatible USB HID multi-touch touchscreens via evdev protocol. It is intended to be a native replacement of hid-multitouch.c driver found in Linux distributions and multimedia/webcamd port. Patch is made for 12-CURRENT and most probably can be applied to recent 11-STABLE and 11.1-RELEASE (not tested) How to test" 1. Apply patch [1] 2. To compile this driver into the kernel, place the following lines into your kernel configuration file: device wmt device usb device evdev Alternatively, to load the driver as a module at boot time, place the following line in loader.conf(5): wmt_load="YES" 3. Install x11-drivers/xf86-input-evdev or x11-drivers/xf86-input-libinput port 4. Tell XOrg to use evdev or libinput driver for the device: ``` Section "ServerLayout" InputDevice "TouchScreen0" "SendCoreEvents" EndSection Section "InputDevice" Identifier "TouchScreen0" Driver "evdev" # Driver "libinput" Option "Device" "/dev/input/eventXXX" EndSection ``` Exact value of "/dev/input/eventXXX" can be obtained with evemu-record utility from devel/evemu. Note1: Currently, driver does not support pens or touchpads. Note2: wmt.ko should be kld-loaded before uhid driver to take precedence over it! Otherwise uhid can be kld-unloaded after loading of wmt. wmt review: https://reviews.freebsd.org/D12017 Raw diff: https://reviews.freebsd.org/D12017.diff *** Beastie Bits BSDMag Programing Languages Infographic (https://bsdmag.org/programm_history/) t2k17 Hackathon Report: Bob Beck on buffer cache tweaks, libressl and pledge progress (http://undeadly.org/cgi?action=article&sid=20170815171854) New FreeBSD Journal (https://www.freebsdfoundation.org/past-issues/resource-control/) NetBSD machines at Open Source Conference 2017 Kyoto (http://mail-index.netbsd.org/netbsd-advocacy/2017/08/10/msg000744.html) *** Feedback/Questions Dan - HDD question (http://dpaste.com/3H6TDJV) Benjamin - scrub of death (http://dpaste.com/10F086V) Jason - Router Opinion (http://dpaste.com/2D9102K) Sohrab - Thanks (http://dpaste.com/1XYYTWF) ***
We read a trip report about FreeBSD in China, look at how Unix deals with Signals, a stats collector in DragonFlyBSD & much more! This episode was brought to you by Headlines Trip Report: FreeBSD in China at COPU and LinuxCon (https://www.freebsdfoundation.org/blog/trip-report-freebsd-in-china-at-copu-and-linuxcon/) This trip report is from Deb Goodkin, the Executive Director of the FreeBSD Foundation. She travelled to China in May 2017 to promote FreeBSD, meet with companies, and participate in discussions around Open Source. > In May of 2017, we were invited to give a talk about FreeBSD at COPU's (China Open Source Promotional Unit) Open Source China, Open Source World Summit, which took place June 21-22, in Beijing. This was a tremendous opportunity to talk about the advantages of FreeBSD to the open source leaders and organizations interested in open source. I was honored to represent the Project and Foundation and give the presentation “FreeBSD Advantages and Applications”. > Since I was already going to be in Beijing, and LinuxCon China was being held right before the COPU event, Microsoft invited me to be part of a women-in-tech panel they were sponsoring. There were six of us on the panel including two from Microsoft, one from the Linux Foundation, one from Accenture of China, and one from Women Who Code. Two of us spoke in English, with everyone else speaking Chinese. It was disappointing that we didn't have translators, because I would have loved hearing everyone's answers. We had excellent questions from the audience at the end. I also had a chance to talk with a journalist from Beijing, where I emphasized how contributing to an open source project, like FreeBSD, is a wonderful way to get experience to boost your resume for a job. > The first day of LinuxCon also happened to be FreeBSD Day. I had my posters with me and was thrilled to have the Honorary Chairman of COPU (also known as the “Father of Open Source in China”) hold one up for a photo op. Unfortunately, I haven't been able to get a copy of that photo for proof (I'm still working on it!). We spent a long time discussing the strengths of FreeBSD. He believes there are many applications in China that could benefit from FreeBSD, especially for embedded devices, university research, and open source education. We had more time throughout the week to discuss FreeBSD in more detail. > Since I was at LinuxCon, I had a chance to meet with people from the Linux Foundation, other open source projects, and some of our donors. With LinuxCon changing its name to Open Source Summit, I discussed how important it is to include minority voices like ours to contribute to improving the open source ecosystem. The people I talked to within the Linux Foundation agreed and suggested that we get someone from the Project to give a talk at the Open Source Summit in Prague this October. Jim Zemlin, the Linux Foundation Executive Director, suggested having a BSD track at the summits. We did miss the call for proposals for that conference, but we need to get people to consider submitting proposals for the Open Source Summits in 2018. > I talked to a CTO from a company that donates to us and he brought up his belief that FreeBSD is much easier to get started on as a contributor. He talked about the steep path in Linux to getting contributions accepted due to having over 10,000 developers and the hierarchy of decision makers, from Linus to his main lieutenants to the layers beneath him. It can take 6 months to get your changes in! > On Tuesday, Kylie and I met with a representative from Huawei, who we've been meeting over the phone with over the past few months. Huawei has a FreeBSD contributor and is looking to add more. We were thrilled to hear they decided to donate this year. We look forward to helping them get up to speed with FreeBSD and collaborate with the Project. > Wednesday marked the beginning of COPU and the reason I flew all the way to Beijing! We started the summit with having a group photo of all the speakers:The honorary chairman, Professor Lu in the front middle. > My presentation was called “FreeBSD Advantages and Applications”. A lot of the material came from Foundation Board President, George-Neville-Neil's presentation, “FreeBSD is not a Linux Distribution”, which is a wonderful introduction to FreeBSD and includes the history of FreeBSD, who uses it and why, and which features stand out. My presentation went well, with Professor Lu and others engaged through the translators. Afterwards, I was invited to a VIP dinner, which I was thrilled about. > The only hitch was that Kylie and I were running a FreeBSD meetup that evening, and both were important! Beijing during rush hour is crazy, even trying to go only a couple of miles is challenging. We made plans that I would go to the meetup and give the same presentation, and then head back to the dinner. Amazingly, it worked out. Check out the rest of her trip report and stay tuned for more news from the region as this is one of the focus areas of the Foundation. *** Unix: Dealing with signals (http://www.networkworld.com/article/3211296/linux/unix-dealing-with-signals.html) Signals on Unix systems are critical to the way processes live and die. This article looks at how they're generated, how they work, and how processes receive or block them On Unix systems, there are several ways to send signals to processes—with a kill command, with a keyboard sequence (like control-C), or through a program Signals are also generated by hardware exceptions such as segmentation faults and illegal instructions, timers and child process termination. But how do you know what signals a process will react to? After all, what a process is programmed to do and able to ignore is another issue. Fortunately, the /proc file system makes information about how processes handle signals (and which they block or ignore) accessible with commands like the one shown below. In this command, we're looking at information related to the login shell for the current user, the "$$" representing the current process. On FreeBSD, you can use procstat -i PID to get that and even more information, and easier to digest form P if signal is pending in the global process queue I if signal delivery disposition is SIGIGN C if signal delivery is to catch it Catching a signal requires that a signal handling function exists in the process to handle a given signal. The SIGKILL (9) and SIGSTOP (#) signals cannot be ignored or caught. For example, if you wanted to tell the kernel that ctrl-C's are to be ignored, you would include something like this in your source code: signal(SIGINT, SIGIGN); To ensure that the default action for a signal is taken, you would do something like this instead: signal(SIGSEGV, SIGDFL); + The article then shows some ways to send signals from the command line, for example to send SIGHUP to a process with pid 1234: kill -HUP 1234 + You can get a list of the different signals by running kill -l On Unix systems, signals are used to send all kinds of information to running processes, and they come from user commands, other processes, and the kernel itself. Through /proc, information about how processes are handling signals is now easily accessible and, with just a little manipulation of the data, easy to understand. links owned by NGZ erroneously marked as on loan (https://smartos.org/bugview/OS-6274) NGZ (Non-Global Zone), is IllumOS speak for their equivalent to a jail > As reported by user brianewell in smartos-live#737, NGZ ip tunnels stopped persisting across zone reboot. This behavior appeared in the 20170202 PI and was not present in previous releases. After much spelunking I determined that this was caused by a regression introduced in commit 33df115 (part of the OS-5363 work). The regression was a one-line change to link_activate() which marks NGZ links as on loan when they are in fact not loaned because the NGZ created and owns the link. “On loan” means the interface belongs to the host (GZ, Global Zone), and has been loaned to the NGZ (Jail) This regression was easy to introduce because of the subtle nature of this code and lack of comments. I'm going to remove the regressive line, add clarifying comments, and also add some asserts. The following is a detailed analysis of the issue, how I debugged it, and why my one-line change caused the regression: To start I verified that PI 20170119 work as expected: booted 20170119 created iptun (named v4sys76) inside of a native NGZ (names sos-zone) performed a reboot of sos-zone zlogin to sos-zone and verify iptun still exists after reboot Then I booted the GZ into PI 20170202 and verified the iptun did not show up booted 20170202 started sos-zone zlogin and verified the iptun was missing At this point I thought I would recreate the iptun and see if I could monitor the zone halt/boot process for the culprit, but instead I received an error from dladm: "object already exists". I didn't expect this. So I used mdb to inspect the dlmgmtd state. Sure enough the iptun exists in dlmgmtd. Okay, so if the link already exists, why doesn't it show up (in either the GZ or the NGZ)? If a link is not marked as active then it won't show up when you query dladm. When booting the zone on 20170119 the llflags for the iptun contained the value 0x3. So the problem is the link is not marked as active on the 20170202 PI. The linkactivate() function is responsible for marking a link as active. I used dtrace to verify this function was called on the 20170202 PI and that the dlmgmtlinkt had the correct llflags value. So the iptun link structure has the correct llflags when linkactivate() returns but when I inspect the same structure with mdb afterwards the value has changed. Sometime after linkactivate() completes some other process changed the llflags value. My next question was: where is linkactivate() called and what comes after it that might affect the llflags? I did another trace and got this stack. The dlmgmtupid() function calls dlmgmtwritedbentry() after linkactivate() and that can change the flags. But dtrace proved the llflags value was still 0x3 after returning from this function. With no obvious questions left I then asked cscope to show me all places where llflags is modified. As I walked through the list I used dtrace to eliminate candidates one at a time -- until I reached dlmgmtdestroycommon(). I would not have expected this function to show up during zone boot but sure enough it was being called somehow, and by someone. Who? Since there is no easy way to track door calls it was at this point I decided to go nuclear and use the dtrace stop action to stop dlmgmtd when it hits dlmgmtdestroycommon(). Then I used mdb -k to inspect the door info for the dlmgmtd threads and look for my culprit. The culprit is doupiptun() caused by the dladm up-iptun call. Using ptree I then realized this was happening as part of the zone boot under the network/iptun svc startup. At this point it was a matter of doing a zlogin to sos-zone and running truss on dladm up-iptun to find the real reason why dladmdestroydatalinkid() is called. So the link is marked as inactive because dladmgetsnapconf() fails with DLADMSTATUSDENIED which is mapped to EACCESS. Looking at the dladmgetsnapconf() code I see the following “The caller is in a non-global zone and the persistent configuration belongs to the global zone.” What this is saying is that if a link is marked "on loan" (meaning it's technically owned/created by the GZ but assigned/loaned to the NGZ) and the zone calling dladmgetsnapconf() is an NGZ then return EACCESS because the configuration of the link is up to the GZ, not the NGZ. This code is correct and should be enforced, but why is it tripping in PI 20170202 and not 20170119? It comes back to my earlier observation that in the 20170202 PI we marked the iptun as "on loan" but not in the older one. Why? Well as it turns out while fixing OS-5363 I fixed what I thought was a bug in linkactivate() When I first read this code it was my understanding that anytime we added a link to a zone's datalink list, by calling zoneadddatalink(), that link was then considered "on loan". My understanding was incorrect. The linkactivate() code has a subtleness that eluded me. There are two cases in linkactivate(): 1. The link is under an NGZ's datalink list but it's lllinkid doesn't reflect that (e.g., the link is found under zoneid 3 but lllinkid is 0). In this case the link is owned by the GZ but is being loaned to an NGZ and the link state should be updated accordingly. We get in this situation when dlmgmtd is restated for some reason (it must resync it's in-memory state with the state of the system). 2. The link is NOT under any NGZ's (zonecheckdatalink() is only concerned with NGZs) datalink list but its llzoneid holds the value of an NGZ. This indicates that the link is owned by an NGZ but for whatever reason is not currently under the NGZ's datalink list (e.g., because we are booting the zone and we now need to assign the link to its list). So the fix is to revert that one line change as well as add some clarifying comments and also some asserts to prevent further confusion in the future. + A nice breakdown by Ryan Zezeski of how he accidently introduced a regression, and how he tracked it down using dtrace and mdb New experimental statistics collector in master (http://dpaste.com/2YP0X9C) Master now has an in-kernel statistics collector which is enabled by default, and a (still primitive) user land program to access it. This recorder samples the state of the machine once every 10 seconds and records it in a large FIFO, all in-kernel. The FIFO typically contains 8192 entries, or around the last 23 hours worth of data. Statistics recorded include current load, user/sys/idle cpu use, swap use, VM fault rate, VM memory statistics, and counters for syscalls, path lookups, and various interrupt types. A few more useful counters will probably be added... I'd like to tie cpu temperature, fork rate, and exec rate in at some point, as well as network and disk traffic. The statistics gathering takes essentially no real overhead and is always on, so any user at the spur of the moment with no prior intent can query the last 23 hours worth of data. There is a user frontend to the data called 'kcollect' (its tied into the buildworld now). Currently still primitive. Ultimately my intention is to integrate it with a dbm database for long-term statistical data retention (if desired) using an occasional (like once-an-hour) cron-job to soak up anything new, with plenty of wiggle room due to the amount of time the kernel keeps itself. This is better and less invasive than having a userland statistics gathering script running every few minutes from cron and has the advantage of giving you a lot of data on the spur of the moment without having to ask for it before-hand. If you have gnuplot installed (pkg install gnuplot), kcollect can generate some useful graphs based on the in-kernel data. Well, it will be boring if the machine isn't doing anything :-). There are options to use gnuplot to generate a plot window in X or a .jpg or .png file, and other options to set the width and height and such. At the moment the gnuplot output uses a subset of statically defined fields to plot but ultimately the field list it uses will be specifiable. Sample image generated during a synth run (http://apollo.backplane.com/DFlyMisc/kcollect03.jpg) News Roundup openbsd changes of note 626 (https://www.tedunangst.com/flak/post/openbsd-changes-of-note-626) Hackerthon is imminent. There are two signals one can receive after accessing invalid memory, SIGBUS and SIGSEGV. Nobody seems to know what the difference is or should be, although some theories have been unearthed. Make some attempt to be slightly more consistent and predictable in OpenBSD. Introduces jiffies in an effort to appease our penguin oppressors. Clarify that IP.OF.UPSTREAM.RESOLVER is not actually the hostname of a server you can use. Switch acpibat to use _BIX before _BIF, which means you might see discharge cycle counts, too. Assorted clang compatibility. clang uses -Oz to mean optimize for size and -Os for something else, so make gcc accept -Oz so all makefiles can be the same. Adjust some hardlinks. Make sure we build gcc with gcc. The SSLcheckprivate_key function is a lie. Switch the amd64 and i386 compiler to clang and see what happens. We are moving towards using wscons (wstpad) as the driver for touchpads. Dancing with the stars, er, NET_LOCK(). clang emits lots of warnings. Fix some of them. Turn off a bunch of clang builtins because we have a strong preference that code use our libc versions. Some other changes because clang is not gcc. Among other curiosities, static variables in the special .openbsd.randomdata are sometimes assumed to be all zero, leading the clang optimizer to eliminate reads of such variables. Some more pledge rules for sed. If the script doesn't require opening new files, don't let it. Backport a bajillion fixes to stable. Release errata. RFC 1885 was obsoleted nearly 20 years ago by RFC 2463 which was obsoleted over 10 years ago by RFC 4443. We are probably not going back. Update libexpat to 2.2.3. vmm: support more than 3855MB guest memory. Merge libdrm 2.4.82. Disable SSE optimizations on i386/amd64 for SlowBcopy. It is supposed to be slow. Prevents crashes when talking to memory mapped video memory in a hypervisor. The $25 “FREEDOM Laptop!” (https://functionallyparanoid.com/2017/08/08/the-25-freedom-laptop/) Time to get back to the original intent of this blog – talking about my paranoid obsession with information security! So break out your tinfoil hats my friends because this will be a fun ride. I'm looking for the most open source / freedom respecting portable computing experience I can possibly find and I'm going to document my work in real-time so you will get to experience the ups (and possibly the downs) of that path through the universe. With that said, let's get rolling. When I built my OpenBSD router using the APU2 board, I discovered that there are some amd64 systems that use open source BIOS. This one used Coreboot and after some investigation I discovered that there was an even more paranoid open source BIOS called Libreboot out there. That started to feel like it might scratch my itch. Well, after playing around with some lower-powered systems like my APU2 board, my Thinkpad x230 and my SPARC64 boxes, I thought, if it runs amd64 code and I can run an open source operating system on it, the thing should be powerful enough for me to do most (if not all) of what I need it to do. At this point, I started looking for a viable machine. From a performance perspective, it looked like the Thinkpad x200, T400, T500 and W500 were all viable candidates. After paying attention on eBay for a while, I saw something that was either going to be a sweet deal, or a throwaway piece of garbage! I found a listing for a Thinkpad T500 that said it didn't come with a power adapter and was 100% untested. From looking at the photos, it seemed like there was nothing that had been molested about it. Obviously, nobody was jumping on something this risky so I thought, “what the heck” and dropped a bit at the opening price of $24.99. Well, guess what. I won the auction. Now to see what I got. When the laptop showed up, I discovered it was minus its hard drive (but the outside plastic cover was still in place). I plugged in my x230's power adapter and hit the button. I got lights and was dropped to the BIOS screen. To my eternal joy, I discovered that the machine I had purchased for $25 was 100% functional and included the T9400 2.54 GHz Core 2 Duo CPU and the 1680×1050 display panel. W00t! First things first, I need to get this machine a hard drive and get the RAM upgraded from the 2GB that it showed up with to 8GB. Good news is that these two purchases only totaled $50 for the pair. An aftermarket 9-cell replacement battery was another $20. Throw in a supported WiFi card that doesn't require a non-free blob from Libreboot at $5.99 off of eBay and $5 for a hard drive caddy and I'm looking at about $65 in additional parts bringing the total cost of the laptop, fully loaded up at just over $100. Not bad at all… Once all of the parts arrived and were installed, now for the fun part. Disassembling the entire thing down to the motherboard so we can re-flash the BIOS with Libreboot. The guide looks particularly challenging for this but hey, I have a nice set of screwdrivers from iFixit and a remarkable lack of fear when it comes to disassembling things. Should be fun! Well, fun didn't even come close. I wish I had shot some pictures along the way because at one point I had a heap of parts in one corner of my “workbench” (the dining room table) and just the bare motherboard, minus the CPU sitting in front of me. With the help of a clip and a bunch of whoops wires (patch cables), I connected my Beaglebone Black to the BIOS chip on the bare motherboard and attempted to read the chip. #fail I figured out after doing some more digging that you need to use the connector on the left side of the BBB if you hold it with the power connector facing away from you. In addition, you should probably read the entire process through instead of stopping at the exciting pinout connector diagram because I missed the bit about the 3.3v power supply need to have ground connected to pin 2 of the BIOS chip. Speaking of that infamous 3.3v power supply, I managed to bend a paperclip into a U shape and jam it into the connector of an old ATX power supply I had in a closet and source power from that. I felt like MacGyver for that one! I was able to successfully read the original Thinkpad BIOS and then flash the Libreboot + Grub2 VESA framebuffer image onto the laptop! I gulped loudly and started the reassembly process. Other than having some cable routing difficulties because the replacement WiFi card didn't have a 5Ghz antenna, it all went back together. Now for the moment of truth! I hit the power button and everything worked!!! At this point I happily scurried to download the latest snapshot of OpenBSD – current and install it. Well, things got a little weird here. Looks like I have to use GRUB to boot this machine now and GRUB won't boot an OpenBSD machine with Full Disk Encryption. That was a bit of a bummer for me. I tilted against that windmill for several days and then finally admitted defeat. So now what to do? Install Arch? Well, here's where I think the crazy caught up to me. I decided to be an utter sell out and install Ubuntu Gnome Edition 17.04 (since that will be the default DE going forward) with full disk encryption. I figured I could have fun playing around in a foreign land and try to harden the heck out of that operating system. I called Ubuntu “grandma's Linux” because a friend of mine installed it on his mom's laptop for her but I figured what the heck – let's see how the other half live! At this point, while I didn't have what I originally set out to do – build a laptop with Libreboot and OpenBSD, I did have a nice compromise that is as well hardened as I can possibly make it and very functional in terms of being able to do what I need to do on a day to day basis. Do I wish it was more portable? Of course. This thing is like a six or seven pounder. However, I feel much more secure in knowing that the vast majority of the code running on this machine is open source and has all the eyes of the community on it, versus something that comes from a vendor that we cannot inspect. My hope is that someone with the talent (unfortunately I lack those skills) takes an interest in getting FDE working with Libreboot on OpenBSD and I will most happily nuke and repave this “ancient of days” machine to run that! FreeBSD Programmers Report Ryzen SMT Bug That Hangs Or Resets Machines (https://hothardware.com/news/freebsd-programmers-report-ryzen-smt-bug-that-hangs-or-resets-machines) It's starting to look like there's an inherent bug with AMD's Zen-based chips that is causing issues on Unix-based operating systems, with both Linux and FreeBSD confirmed. The bug doesn't just affect Ryzen desktop chips, but also AMD's enterprise EPYC chips. It seems safe to assume that Threadripper will bundle it in, as well. It's not entirely clear what is causing the issue, but it's related to the CPU being maxed out in operations, thus causing data to get shifted around in memory, ultimately resulting in unstable software. If the bug is exercised a certain way, it can even cause machines to reset. The revelation about the issue on FreeBSD was posted to the official repository, where the issue is said to happen when threads can lock up, and then cause the system to become unstable. Getting rid of the issue seems as simple as disabling SMT, but that would then negate the benefits provided by having so many threads at-the-ready. On the Linux side of the Unix fence, Phoronix reports on similar issues, where stressing Zen chips with intensive benchmarks can cause one segmentation fault after another. The issue is so profound, that Phoronix Test Suite developer Michael Larabel introduced a special test that can be run to act as a bit of a proof-of-concept. To test another way, PTS can be run with this command: PTS_CONCURRENT_TEST_RUNS=4 TOTAL_LOOP_TIME=60 phoronix-test-suite stress-run build-linux-kernel build-php build-apache build-imagemagick Running this command will compile four different software projects at once, over and over, for an hour. Before long, segfaults should begin to appear (as seen in the shot above). It's not entirely clear if both sets of issues here are related, but seeing as both involve stressing the CPU to its limit, it seems likely. Whether or not this could be patched on a kernel or EFI level is something yet to be seen. TrueOS - UNSTABLE update: 8/7/17 (https://www.trueos.org/blog/unstable-update-8717/) A new UNSTABLE update for TrueOS is available! Released regularly, UNSTABLE updates are the full “rolling release” of TrueOS. UNSTABLE includes experimental features, bugfixes, and other CURRENT FreeBSD work. It is meant to be used by those users interested in using the latest TrueOS and FreeBSD developments to help test and improve these projects. WARNING: UNSTABLE updates are released primarily for TrueOS and FreeBSD testing/experimentation purposes. Update and run UNSTABLE “at your own risk”. Note: There was a CDN issue over the weekend that caused issues for early updaters. Everything appears to be resolved and the update is fully available again. If you encountered instability or package issues from updating on 8/6 or 8/5, roll back to a previous boot environment and run the update again. Changes: UNSTABLE .iso and .img files beginning with TrueOS-2017-08-3-x64 will be available to download from http://download.trueos.org/unstable/amd64/. Due to CDN issues, these are not quite available, look for them later today or tomorrow (8/8/17). This update resyncs all ports with FreeBSD as of 8.1.2017. This includes: New/updated FreeBSD Kernel and World & New DRM (Direct Rendering Manager) next. Experimental patch for libhyve-remote: (From htps://github.com/trueos/freebsd/commit/a67a73e49538448629ea27, thanks araujobsd) The libhyve-remote aims to abstract functionalities from other third party libraries like libvncserver, freerdp, and spice to be used in hypervisor implementation. With a basic data structure it is easy to implement any remote desktop protocol without digging into the protocol specification or third part libraries – check some of our examples.We don't statically link any third party library, instead we use a dynamic linker and load only the functionality necessary to launch the service.Our target is to abstract functionalities from libvncserver, freerdp and spice. Right now, libhyve-remote only supports libvncserver. It is possible to launch a VNC server with different screen resolution as well as with authentication.With this patch we implement support for bhyve to use libhyve-remote that basically abstract some functionalities from libvncserver. We can: Enable wait state, Enable authentication, Enable different resolutions< Have a better compression. Also, we add a new -s flag for vncserver, if the libhyve-remote library is not present in the system, we fallback to bhyve RFB implementation. For example: -s 2,fbuf,tcp=0.0.0.0:5937,w=800,h=600,password=1234567,vncserver,wait New SysAdm Client pages under the System Management category: System Control: This is an interface to browse all the sysctl's on the system. Devices: This lists all known information about devices on the designated system. Lumina Theming: Lumina is testing new theming functionality! By default (in UNSTABLE), a heavily customized version of the Qt5ct engine is included and enabled. This is intended to allow users to quickly adjust themes/icon packs without needing to log out and back in. This also fixes a bug in Insight with different icons loading for the side and primary windows. Look for more information about this new functionality to be discussed on the Lumina Website. Update to Iridium Web Browser: Iridium is a Chromium based browser built with user privacy and security as the primary concern, but still maintaining the speed and usability of Chromium. It is now up to date – give it a try and let us know what you think (search for iridium-browser in AppCafe). Beastie Bits GhostBSD 11.1 Alpha1 is ready (http://www.ghostbsd.org/11.1-ALPHA1) A Special CharmBUG announcement (https://www.meetup.com/CharmBUG/events/242563414/) Byhve Obfuscation Part 1 of Many (https://github.com/HardenedBSD/hardenedBSD/commit/59eabffdca53275086493836f732f24195f3a91d) New BSDMag is out (https://bsdmag.org/download/bsd-magazine-overriding-libc-functions/) git: kernel - Lower VMMAXUSER_ADDRESS to finalize work-around for Ryzen bug (http://lists.dragonflybsd.org/pipermail/commits/2017-August/626190.html) Ken Thompson corrects one of his biggest regrets (https://twitter.com/_rsc/status/897555509141794817) *** Feedback/Questions Hans - zxfer (http://dpaste.com/2SQYQV2) Harza - Google Summer of Code (http://dpaste.com/2175GEB) tadslot - Microphones, Proprietary software, and feedback (http://dpaste.com/154MY1H) Florian - ZFS/Jail (http://dpaste.com/2V9VFAC) Modifying a ZFS root system to a beadm layout (http://dan.langille.org/2015/03/11/modifying-a-zfs-root-system-to-a-beadm-layout/) ***
© 2020-2025 Ivy Podcast Discovery LLC
