Dissecting Nvidia Blackwell - Tensor Cores, PTX Instructions, SASS, Floorsweep, Yield

The custom AI ASIC state of play (May 2026) — Broadcom deals, Google TPUs, Meta MTIA & beyond

## Episode Summary In this episode, we cover: - **The custom AI ASIC state of play (May 2026) — Broadcom deals, Google TPUs, Meta MTIA & beyond** (toms_hardware) - [Read more](https://www.tomshardware.com/tech-industry/semiconductors/custom-ai-asics-examined-from-broadcom-to-mtia) - **Large-scale, SRAM-based LLM Inference Deployment (Groq)** (semiengineering) - [Read more](https://semiengineering.com/large-scale-sram-based-llm-inference-deployment-groq/) - **Taiwan raids 12 locations in its first formal crackdown on Nvidia AI chip smuggling — hunts three fugitives for document forgery, fraudulent declarations in Super Micro smuggling case** (toms_hardware) - [Read more](https://www.tomshardware.com/desktops/servers/taiwan-raids-12-locations-in-its-first-formal-crackdown-on-nvidia-ai-chip-smuggling-hunts-three-fugitives-for-document-forgery-fraudulent-declarations-in-super-micro-smuggling-case) - **Scaling the AI-Ready Data Center with NVIDIA RTX PRO 4500 Blackwell Server Edition and NVIDIA vGPU 20** (nvidia_dev) - [Read more](https://developer.nvidia.com/blog/scaling-the-ai-ready-data-center-with-nvidia-rtx-pro-4500-blackwell-server-edition-and-nvidia-vgpu-20/) - **Nvidia to launch AI chip for data centers in space - Yahoo Finance** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMinAFBVV95cUxOVl94UVFMaFhJUmM5aWM1QlRRajRVTHU0SG9ZNG1vMU8xVk55LTRtdDFFZFlXREd3akRwYThWVVk1WFdwTmNnRzd2RWNXR3JJRUVhck1BOHk5b19jSXZnSEVsYlF2RmZhdHdReEotZ1ItWkkteDRENUI5Sy1TNTZocWJsLWlVNUFJZ1B5dnYzYkh5YmJESTR1S2U2dkw?oc=5) --- *Sponsored by LimitLess AI*

AMD Intros Instinct MI350P Accelerator: CDNA 4 Comes to PCIe Cards

## Episode Summary In this episode, we cover: - **AMD Intros Instinct MI350P Accelerator: CDNA 4 Comes to PCIe Cards** (servethehome) - [Read more](https://www.servethehome.com/amd-intros-instinct-mi350p-accelerator-cdna-4-comes-to-pcie-cards/) - **Huawei could seize China’s AI chip crown in 2026 as Nvidia's H200 shipments stall in regulatory limbo — Beijing pushes homegrown AI hardware dominance in a market projected to hit $67 billion by 2030 - Tom's Hardware** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMi_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?oc=5) - **Meta expands Nvidia deal to use millions of AI chips in data center build-out, including standalone CPUs - CNBC** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMif0FVX3lxTFBiLXZYX1ZjaTRaenZCcmpJT0FZTksxVFpNOWhYOXBHenZDUjlKLXBoTDJuOHluYzRiSWx0YVV6Slljb3NBem9CLVlJUHlTR1FIQk15ODZGRm4yaXduckZ6YW9DNTJoYkZveDFwS3Y1amFOTkRiSi1lNDYxZlQwMFXSAYQBQVVfeXFMTVBJeVBlbFhKZDhpWE56QTZ4VDFJYjlrSk50eFgtQzViYUduMHRZRmJ2amNRdjZDeVFmUnpfNE54LUhqR2ZTSWR0eXBDcE13Q2wxekszejRCTFlKSHRZWDlWZDgzb0pqZjE4QTdrd0FhZDhQSTNHWDZDMlQtNUYya1RTVkJO?oc=5) - **Nvidia GPUs to Google TPUs: Breaking down all the AI chips - CNBC** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMiqgFBVV95cUxQUU9vT2lsNmEyRTdfdnBFSkVJWFlzVnlFbmY4MEdTZ2VtR3RweU5KakQzczkwOENWWlZfaVp1NEdENzFSSXJuRU1Fa0drcmVJa3JBVms3SU1iZzNDeEFrLWROTEFVUnQyTFk4blNvVlBxbzJVM0dodnl4WTRaUGhFYXdfaUJlM1JDeWhOSi1GQ2VHaUdQVmllYkR2cElkZF84cFFKc25pVlNLUQ?oc=5) - **AMD announces MI350P PCIe AI accelerator card with 144GB of HBM3E — roughly 40% faster in FP16 and FP8 theoretical compute compared to Nvidia's H200 NVL competitor** (toms_hardware) - [Read more](https://www.tomshardware.com/pc-components/gpus/amd-announces-mi350p-pcie-ai-accelerator-card-with-144gb-of-hbm3e-roughly-40-percent-faster-in-fp16-and-fp8-theoretical-compute-compared-to-nvidias-h200-nvl-competitor) --- *Sponsored by LimitLess AI*

Dissecting Nvidia Blackwell - Tensor Cores, PTX Instructions, SASS, Floorsweep, Yield

## Episode Summary In this episode, we cover: - **Dissecting Nvidia Blackwell - Tensor Cores, PTX Instructions, SASS, Floorsweep, Yield** (semianalysis) - [Read more](https://newsletter.semianalysis.com/p/dissecting-nvidia-blackwell-tensor) - **NVIDIA Dominance Unshaken as Data Centers Hesitate on Google TPUs - 조선일보** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMiiAFBVV95cUxOR3VEMkZzSVFaNUlvaWc0czJzUkVDYlRuSGd2bzJvTlFoa3VuVmZscFNzYjRERzZpVVJTY1lrT0hYYWdJQ2RocHNuYXdSZ2xfNHp1Wi1qa3M1WWptU2Zqdk1ST3ZBYk5SNE84ZFJwY1BOZFdtQl95NXg0Z1ZvQlFESjZwWTVHTFg5?oc=5) - **Google Expands TPU Sales Amid Nvidia's AI Chip Dominance - 조선일보** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMiiAFBVV95cUxPQjNPTFZCRFFsQm9jcVJYaFVkb1I0TGdmU3FaYjNJX1JTOVNONGhTb1J0VFNrdHc3N0F5emR4MUt2T1pNREFyRDgxZHdDTEdMTFMtQ3pJdUlvSGJXVHF6TzFfSXNWM28zR1FYLUstVFNNOE5MTDlUMndIWERNUDVVc004LUNiakxj?oc=5) - **E2E Networks acquires 1,024 Nvidia B200 GPU cluster for Chennai data center - Data Center Dynamics** (google_nvidia) - [Read more](https://news.google.com/rss/articles/CBMiugFBVV95cUxOR29qZVM5SkFFdjBIUmlQcTNFejk4eHJxVklNMVo1d18yRmZpTUtHdGN1Zk95V0hEN3VieGk1LW52dHB4TzhrUzdNZFFONU9hWnYycEN6ckc1SXZLcVU5OEw4OUp3TC0zT0trZUhYSGI0X29MTkd0WllnSXR6eFpkbnd1Q081OEgxeERMLTdLeVJ2MWR4bW50MHBrNGZKbnR1QjlqOUY2RUQ1T3h5bFFhcDNEZl9LbmM1aUE?oc=5) - **OpenAI Got The Whole AI Squad To Accelerate Large-Scale AI Training – AMD, NVIDIA, Intel, Microsoft & Broadcom All-In On MRC** (wccftech) - [Read more](https://wccftech.com/openai-accelerates-large-scale-ai-training-amd-nvidia-intel-microsoft-broadcom-mrc/) --- *Sponsored by LimitLess AI*

AI chipmaker Cerebras targets $3.5 billion raise in IPO - The Tech Buzz

## Episode Summary In this episode, we cover: - **AI chipmaker Cerebras targets $3.5 billion raise in IPO - The Tech Buzz** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMikAFBVV95cUxNdldyc1ZQRENaNW16WmtGVHNNVGo5cENmM2NEa3g2NEFacnRfWXZLankxNGJOR1JuZmtiQUpHMnRCNy1MNHJYeVdFN0sxbHhidkFhOVhreUxnT2JTVWdMZUtOT1d3VUhieDMwNkFwS05hNWNSZzJrZVYxd01tYUdubWFzM2lZMWdWTU1FVVpkbGM?oc=5) - **Nvidia CEO Jensen Huang says China should not have Blackwell or Rubin AI GPUs — firmly states US should have 'the first, the most, and the best' when it comes to AI hardware** (toms_hardware) - [Read more](https://www.tomshardware.com/tech-industry/artificial-intelligence/nvidia-ceo-jensen-huang-says-china-should-not-have-blackwell-or-rubin-ai-gpus-firmly-states-us-should-have-the-first-the-most-and-the-best-when-it-comes-to-ai-hardware) - **NVIDIA Leads in Data Center GPU Market: Will Blackwell Keep It Ahead? - Yahoo Finance** (google_nvidia) - [Read more](https://news.google.com/rss/articles/CBMif0FVX3lxTE45TU1qajRialRGMU1qZTlGMWVQSFlsUk1STlFiZ0o5ZnBiT1hMdWZwZGNhR190RWxBREx2bXZGUXNONVMwQUxlMTZuaHZtZUZLT3lkYlhhMGN2U2V3ZzJnTGhuaGR1Xzd6MWJtSTRyU2p1a1QwYVZMb1FDWEszZVE?oc=5) - **Cerebras eyes US$40 billion IPO in high-stakes AI chip challenge to Nvidia - digitimes** (google_gpu) - [Read more](https://news.google.com/rss/articles/CBMimAFBVV95cUxOeVd0dldRVlZSSTJoOWNzbURvcXFnbjRiVGVIV3p2dTZwYjMwcU9sWGU1azJmbzR0dTVKaDVWd3dfNy1Va0o3YU9SRnlNTmFaOHVTU0p4T2FJdXFzWWhVMHdzZGlDUDF1dWlETzJTQS1uLWlHMWQ1dVlUTVRTYVE2RnpFcmV1TmtjMnF3NlZOSV9ibE54am9lOA?oc=5) - **NVIDIA’s CEO Jensen Huang Draws a Hard Line on China: No Blackwell, No Rubin, But US Firms Must Still Fight In Global Markets** (wccftech) - [Read more](https://wccftech.com/nvidia-ceo-jensen-huang-draws-a-hard-line-on-china-no-blackwell-no-rubin/) --- *Sponsored by LimitLess AI*

NVIDIA Feynman GPUs Push Power Semi Content To $191,000, 17 Times Increase Over Blackwell As Industry Embraces 800V DC Architectures

## Episode Summary In this episode, we cover: - **NVIDIA Feynman GPUs Push Power Semi Content To $191,000, 17 Times Increase Over Blackwell As Industry Embraces 800V DC Architectures** (wccftech) - [Read more](https://wccftech.com/nvidia-feynman-gpus-push-power-semi-content-17-times-higher-vs-blackwell/) - **Analyst Warns Google’s Intel TPU Bet Hinges on a Brutal Yield Jump From 90% to 98%** (wccftech) - [Read more](https://wccftech.com/analyst-warns-googles-intel-tpu-bet-hinges-on-a-brutal-yield-jump-from-90-to-98/) - **Bain Capital's data center unit removes disgraced tenant suspected of smuggling Nvidia GPUs to China — Megaspeed previously alleged to have spent roughly $2 billion on AI processors for illicit distribution - Tom's Hardware** (google_nvidia) - [Read more](https://news.google.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?oc=5) - **AI chipmaker Cerebras targets up to $4 billion in IPO - ET Telecom** (google_ai_chip) - [Read more](https://news.google.com/rss/articles/CBMivAFBVV95cUxOSTd4R0ZiaWVTZERycUctUW9mZExFMkcxZjA0OThhYXRBejN2TnVhbThfWnZsWFBVUDFqVVZMZjBKVnZaUElDOFRISGlGWXVoTmdTa0NjRnZ4VDdsbE9icGNaWDlCTTJTajc2ZUJvZk1mb2VxOVRZcWpYUlNqaGpKa2dvMHE0eWs0SmtrUFRXQ3RtMVNBV0hrS2xxRHZZczUwcXppMlZPemNMV01XTkVPUjNRdVNSVVdOTExjddIBwgFBVV95cUxNV3h3RXUyZGxueXM5YUladng5Q2ZGNHhySG9KTU9jeTJzSXlEdmZNcnc2UEVyUFdCaUNXV0kwRVZ3ZDVwUUN2cXhMS05CWWdldnhmVnVtMkxhTm05RndzSjk5ZU5RUjRma084UlNCaFNSWmh3VkNCR3R6cUl3Um5TbHhkcEdYaE9xNGVabDczaWZ5dHNBYUpHUmlENjByT1ZhbWhGakZlRXc0QUZpb1VQNGpQWWhCMGRBWlRmdUxiVWVQQQ?oc=5) - **InferenceX v2: NVIDIA Blackwell Vs AMD vs Hopper - Formerly InferenceMAX** (semianalysis) - [Read more](https://newsletter.semianalysis.com/p/inferencex-v2-nvidia-blackwell-vs) --- *Sponsored by LimitLess AI*