AI Chips news and analysis on Web Pulse: 6079 curated articles tracking the latest AI Chips developments, tools, and research, updated continuously from vetted sources.
AI News Digest for August 26 covers 30 stories, including UK firms gaining early access to five million combat images, Nvidia's Groq 3 LPX claiming speed improvements, and China shipping nearly 90% of the world's bipedal…
OpenAI unveiled benchmarks for its custom inference ASIC Jalapeño, co-designed with Broadcom, claiming 1.5–1.9x better performance per watt and up to 3.6x lower latency versus Nvidia's GB200/GB300 on several large models…
Nvidia has begun mass production of its Groq 3 LPX inference accelerator, with Samsung Electronics manufacturing the key chips on its 4-nanometer process, raising expectations that Samsung's loss-making foundry business …
Researchers presented the Transformer Accelerator (TFA), a synthesizable INT8 hardware chip for transformer inference, achieving zero mismatches across 25 tests and 34 constrained-random runs, with 100% functional covera…
OpenAI and Broadcom's Jalapeño inference chip, unveiled Tuesday at the Hot Chips conference, delivers 1.5 to 1.9 times more work per watt and 1.7 to 3.6 times lower latency on tested models, according to OpenAI. The chip…
Asian equities paused on Wednesday as sliding oil prices and falling bond yields signaled cautious optimism over the Strait of Hormuz, while traders braced for Nvidia's second-quarter earnings. Brent crude futures fell m…
Tesla and SpaceX unveiled Terafab, a chip factory designed for both Earth and orbit, aiming to produce AI chips at terawatt scale using solar power. The facility will employ 1 billion Tesla Optimus robots and launch 10 m…
Raymond James upgraded Advanced Micro Devices (AMD) to Strong Buy on August 25 and raised its price target from $565 to $641, citing a server CPU market expected to grow at a 44% compound annual rate to about $201 billio…
OpenAI unveiled Jalapeño, an in-house inference ASIC and system built with Broadcom, at Hot Chips 2026, claiming it outperforms NVIDIA GB200 and GB300 in throughput per kilowatt and latency for OpenAI's inference workloa…
Nvidia's Groq 3 LPX inference rack entered full production on August 24, with shipments expected before the end of 2026, marking the first hardware from its $20 billion licensing deal with Groq. The rack, featuring 256 L…
Cerebras Systems detailed the rack architecture for its CS-4 AI system at the Hot Chips conference on August 25, introducing the Nexus platform that integrates three WSE-3 Turbo processors with modular power, cooling, an…
Google announced its eighth-generation tensor processor units (TPUs), the TPU 8t for training and TPU 8i for inference, at Hot Chips 2026, marking the first time the company has released two TPU generations in a single y…
SambaNova Systems presented technical details of its fifth-generation SN50 reconfigurable dataflow unit (RDU) at Hot Chips 2026, claiming it delivers 5x the FLOPS of the SN40 and scales to 256+ chips. The SN50 uses HBM2e…
Nvidia unveiled its Vera CPU, Rubin GPU, BlueField-4 DPU, and Spectrum-X networking at Hot Chips 2026, claiming the Vera CPU compiles the Linux kernel 14-22% faster than AMD's 96-core EPYC 9655P and delivers a 1.8x impro…
OpenAI revealed benchmark results for its first custom inference chip, Jalapeño, co-developed with Broadcom on TSMC's 3nm node, at Hot Chips 2026 on August 25. The chip delivers 1.5x–1.9x more AI work per watt and up to …
Microsoft unveiled the architecture of its second-generation Maia 200 AI accelerator at Hot Chips 2026, a 3nm chip with 140 billion transistors, 750W TDP, and 10,000 TFLOPS of FP4 performance, designed for deployment in …
OpenAI's 700W Jalapeño ASIC outperforms Nvidia's 1,400W flagship GPU, according to a report from hardware news outlet. The custom chip achieves higher performance while consuming half the power, marking a significant adv…
NVIDIA announced at Hot Chips 2026 that its Vera Rubin racks will incorporate Groq 3 LPU accelerators, purchased from Groq as part of an acquihire, to boost low-latency decode performance for agentic AI workloads. A sing…
Apple Inc. announced its M5 Ultra and M6 chips, set to debut in September in the Mac mini and Mac Studio desktops, featuring 2nm technology and quad-die design to significantly boost local agentic AI performance.
Meta Platforms Inc. presented its Meta Training and Inference Accelerators (MTIA) custom AI silicon at Hot Chips 2026, detailing a roadmap of four generations (MTIA 300, 400, 450, 500) to reduce reliance on commodity GPU…