AI Chips news and analysis on Web Pulse: 6172 curated articles tracking the latest AI Chips developments, tools, and research, updated continuously from vetted sources.
NVIDIA detailed its new "Rubin" GPU architecture, which delivers 50 PetaFLOPS at sparse NVFP4 operations, features 224 streaming multiprocessors, 288 GB of HBM4 memory, 336 billion transistors, and 896 Tensor Cores with …
Gartner named Nvidia the 'company to beat' in AI network fabric, citing its AI accelerator dominance and broad data center networking portfolio, but warned that a shift from training to inference and the rise of agentic …
Alphabet is developing a custom server chip codenamed 'Frozen v2' optimized for its Gemini AI models, targeting a 2028 launch and designed to generate 6 to 10 times more tokens per unit of power than Google's latest Tens…
Google is developing a custom chip designed specifically for its own AI models, a move that could reduce reliance on external suppliers and improve performance. The chip initiative is expected to impact Google's cost str…
Halvledarbolag lyfter Wall Street, med S&P 500 upp 0,9 procent och Nasdaq upp 1,2 procent. Micron rusar 12,9 procent, medan AMD, Marvell och Intel stiger 7,5–7,9 procent. Nvidia klättrar 1,5 procent. Alphabet backar 0,8 …
Google is developing a server chip codenamed Frozen v2 that hardwires part of its Gemini AI architecture into silicon, targeting six to ten times the efficiency of current TPUs and a 2028 deployment, according to a repor…
NVIDIA detailed its custom Armv9.2 "Vera" CPU with 88 cores, 176 threads, and support for up to 1.5 TB of LPDDR5X memory, optimized for data analysis, high-performance single-threaded tasks, agentic AI, and large-scale s…
Nvidia unveiled the Vera CPU architecture, designed as a single-threaded CPU for agentic AI and reinforcement learning workloads, featuring 88 Olympus cores with 176 threads and 1.2 TB/s of LPDDR5X memory bandwidth. The …
Google launched three new Gemini models on Tuesday, including Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, all at its low-cost Flash tier, emphasizing efficiency over raw power. The company claims Gemini 3.6 Fl…
Acer's Predator Atlas 8 gaming handheld, powered by Intel Arc G3 APUs, will start at £999.99 ($1,346) when it launches this autumn, according to a UK product page update. The cheapest model exceeds $1,000, with US pricin…
Nvidia confirmed Monday that its Vera Rubin platform is in full production, with OpenAI planning to deploy the system at scale in the third quarter. Ian Buck, Nvidia's vice president of accelerated computing, told report…
Qualcomm CEO Cristiano Amon, marking his fifth anniversary this summer, says the future of computing is AI and that by definition Qualcomm is an AI company. He predicts AI will be bigger than people expect, comparing it …
NVIDIA launched the Vera Rubin platform, built from chip to grid to deliver the highest performance per watt and lowest token cost, with production ramping at partners CoreWeave, Google Cloud, Microsoft Azure and Oracle …
SK Hynix now generates 65% of its revenue from the United States, driven by AI infrastructure demand for its high-bandwidth memory chips, not crypto mining. The South Korean chipmaker posted 97.1 trillion won ($64.1 bill…
Nvidia launched its Spectrum-6 Ethernet switch for AI factories, with CoreWeave, Microsoft, Nebius, SpaceX, and Tesla among initial adopters. The switch offers high switching capacity to support AI models in large comput…
Nvidia's Spectrum-6 Ethernet switch, delivering 102.4 Tb/s switching capacity on a single chip, is now being installed at major AI factories to manage networks connecting hundreds of thousands of GPUs. The sixth-generati…
AMD will host its Advancing AI 2026 summit on July 22-23 in San Francisco to showcase its MI350 series GPUs, which feature 288 GB of HBM3e memory versus Nvidia's 180 GB, as the chipmaker challenges Nvidia's 80-95% domina…
Nvidia's latest AI GPU designs are now shipping to customers, and the company holds an estimated 80-81% share of the data-center AI GPU market, according to IDC data. The Rubin platform entered full production in early 2…
TSMC posted its fifth consecutive record quarter on July 16, with revenue hitting $40.2 billion (up 36% year-over-year) and net profit surging 77%, as CEO C.C. Wei flagged that agentic AI is creating a second wave of CPU…
Nvidia's next-generation Vera Rubin AI platform is on schedule and already being tested by customers, CEO Jensen Huang confirmed, dismissing rumors of production delays. The platform, integrating the Vera CPU and Rubin G…