Hyperscaler custom ASICs https://ai-dc-parts.pages.dev/parts/compute-silicon-and-packaging/hyperscaler-custom-asics-xpus News and posts about Hyperscaler custom ASICs on AI Data Center Parts Huawei's Ascend 910C sold out in China, but Nvidia's H200 supply gap persists https://www.digitimes.com/news/a20260828PD207/huawei-ascend-nvidia-chips-china-2026.html Fri, 28 Aug 2026 04:45:00 GMT Although Huawei's Ascend 910C is sold out in China, the inability of domestic chips to fully replace Nvidia's H200 suggests ongoing global AI compute constraints. (DigiTimes) Insight: China's next AI frontier is silicon, not smarter models https://www.digitimes.com/news/a20260828PD213/z.ai-performance-chips-language-testing.html Fri, 28 Aug 2026 02:55:16 GMT US export restrictions on advanced AI chips are forcing China's AI companies to shift focus from model performance to the ability to operate models at scale using domestically manufactured chips. (DigiTimes) Analysis: Nvidia beats, Taiwan wins — and hyperscalers race to make AI pay https://www.digitimes.com/news/a20260828PD203/nvidia-taiwan-revenue-capex-google.html Fri, 28 Aug 2026 02:37:54 GMT Nvidia reported massive revenue growth, driven by the global expansion of AI infrastructure and increased purchases from hyperscalers. (DigiTimes) First Benchmarks Revealed for Jalapeño, OpenAI’s Clean-Sheet General Purpose AI Accelerator ASIC https://www.eetimes.com/first-benchmarks-revealed-for-jalapeno-openais-clean-sheet-general-purpose-ai-accelerator-asic Thu, 27 Aug 2026 22:14:16 GMT OpenAI unveiled Jalapeño, a custom, purpose-built ASIC designed from the ground up specifically for AI workloads. (EE Times) OpenAI Details Jalapeño Inference Chip Performance and Multigenerational Roadmap https://www.hpcwire.com/off-the-wire/openai-details-jalapeno-inference-chip-performance-and-multigenerational-roadmap Thu, 27 Aug 2026 20:04:38 GMT OpenAI announced the performance of its custom inference chip, Jalapeño, demonstrating improvements in both throughput and latency. (HPCwire) Hot Chips 2026: OpenAI's Jalapeño AI ASIC unpacked — accelerator developed using AI achieves efficiency and throughput gains against power-hungry Blackwell https://www.tomshardware.com/tech-industry/artificial-intelligence/hot-chips-2026-openais-jalapeno-ai-asic-unpacked-accelerator-developed-using-ai-achieves-efficiency-and-throughput-gains-against-power-hungry-blackwell Thu, 27 Aug 2026 13:00:00 GMT OpenAI unveiled its Jalapeño AI ASIC accelerator, noting it offers good performance per watt and low latency for inference tasks compared to Nvidia's Blackwell. (Tom's Hardware) Meta's new MTIA 400 chip has a split personality: Training AI and serving ads https://www.theregister.com/systems/2026/08/26/metas-new-mtia-400-chip-has-a-split-personality-training-ai-and-serving-ads/5292727 Wed, 26 Aug 2026 21:18:04 GMT Meta introduced the MTIA 400 chip, an AI accelerator that offers speed improvements over Blackwell but has not yet replaced competitive offerings from AMD or Nvidia. (The Register) Accelerating Silicon Design for Physical AI https://www.eetimes.com/accelerating-silicon-design-for-physical-ai Wed, 26 Aug 2026 12:00:00 GMT Physical AI mandates complex silicon designs that require multi-vendor supply chains and full lifecycle security. (EE Times) OpenAI Jalapeno Custom AI ASIC at Hot Chips 2026 https://www.servethehome.com/openai-jalapeno-asic-at-hot-chips-2026 Wed, 26 Aug 2026 00:45:45 GMT OpenAI presented its custom AI accelerator, Jalapeño, at Hot Chips 2026. (ServeTheHome) Google’s TPUv8s for Training and Inference at Hot Chips 2026 https://www.servethehome.com/googles-tpuv8s-for-training-and-inference-at-hot-chips-2026 Wed, 26 Aug 2026 00:15:29 GMT Google unveiled its new eighth-generation TPU family (TPU 8t for training and TPU 8i for inference) at Hot Chips 2026. (ServeTheHome) SambaNova’s SN50 RDU for AI at Hot Chips 2026 https://www.servethehome.com/sambanovas-sn50-rdu-for-ai-at-hot-chips-2026 Tue, 25 Aug 2026 23:45:32 GMT SambaNova showcased its latest-generation RDU, the SN50, at the Hot Chips conference. (ServeTheHome) Microsoft’s Maia 200 AI Accelerator at Hot Chips 2026 https://www.servethehome.com/microsofts-maia-200-accelerator-at-hot-chips-2026 Tue, 25 Aug 2026 22:45:29 GMT Microsoft unveiled details about Maia 200, a second-generation AI accelerator designed for server inference, at the Hot Chips 2026 conference. (ServeTheHome) OpenAI’s 700W Jalapeño ASIC outpaces 1,400W Nvidia flagship GPU — claims up to 1.9x throughput per kilowatt and 3.6x lower latency, co-developed with Broadcom https://www.tomshardware.com/tech-industry/semiconductors/openai-says-its-jalapeno-chip-beats-nvidias-gb300-in-first-published-benchmarks Tue, 25 Aug 2026 18:05:23 GMT OpenAI debuted an in-house chip, claiming it offers superior performance and efficiency compared to Nvidia's high-end GPUs. (Tom's Hardware) AMD Acquires Taalas to Advance Compute Solutions for Rapidly Growing AI Inference Market https://ir.amd.com/news-events/press-releases/detail/1296/amd-acquires-taalas-to-advance-compute-solutions-for-rapidly-growing-ai-inference-market Thu, 06 Aug 2026 20:05:00 GMT AMD acquired Taalas to enhance compute solutions specifically for the fast-growing AI inference market. (newsroom:amd) AMD and Cerebras Announce Industry-Leading Ultra-Low-Latency and High Throughput AI Inference Solution https://ir.amd.com/news-events/press-releases/detail/1293/amd-and-cerebras-announce-industry-leading-ultra-low-latency-and-high-throughput-ai-inference-solution Thu, 23 Jul 2026 17:45:00 GMT AMD and Cerebras jointly announced a new AI inference solution that promises both ultra-low latency and high throughput. (newsroom:amd) Amazon’s AI Resurgence: AWS & Anthropic’s Multi-Gigawatt Trainium Expansion https://semianalysis.com/2025/09/03/amazons-ai-resurgence-aws-anthropics-multi-gigawatt-trainium-expansion Wed, 03 Sep 2025 20:55:46 GMT The article discusses Amazon's AWS and Anthropic's efforts to build out multi-gigawatt AI infrastructure, noting that while AWS dominates cloud profits, its transition to the new XPU/GPU era presents challenges. (SemiAnalysis)