AI Chips news and analysis on Web Pulse: 6203 curated articles tracking the latest AI Chips developments, tools, and research, updated continuously from vetted sources.
OpenAI previewed its GPT-5.6 model family on June 26, 2026, with the flagship Sol variant targeting 750 tokens per second via a partnership with Cerebras Systems. Cerebras' wafer-scale compute could outperform convention…
Tensordyne announced its Napier inference processor, which uses logarithmic math to convert matrix multiplications into additions, and said the 3nm TSMC chip is now in production. The company claims the approach can deli…
Positron AI is in talks to raise $750 million across two funding rounds to challenge Nvidia's dominance in AI inference chips. The Reno-based startup claims its Atlas system delivers 3.5x better performance-per-watt than…
Bank of America reiterated its Buy rating on Nvidia stock with a $350 price target, calling it a compelling buying opportunity at a seven-year low valuation. Analyst Vivek Arya argued that the market has overcorrected on…
Open interest in SK Hynix perpetual futures on Trade.xyz surged 210% over the past month, driven by institutional appetite and AI-fueled demand ahead of the company's ADR listing. The decentralized platform now holds bet…
Rambus Inc. announced its DDR5 9600 Client Memory Module Chipset for AI PCs, featuring a new Gen 2 Client Clock Driver that supports up to 9600 MT/s. The chipset addresses signal degradation and timing issues in high-spe…
Grid interconnection delays, not electricity shortages, are slowing AI data center construction, with projects like Stargate in Abilene, Texas, facing multi-year waits for utility studies and transmission upgrades. Media…
SambaNova combined four Nvidia H200 GPUs with 16 SN50 RDU chips to achieve 763 tokens per second on MiniMax M2.7 at a 10,000-token context, separating prefill and decode stages to accelerate inference. The benchmark, rep…
Intel-backed AI chip startup SambaNova released third-party benchmarks showing its heterogeneous compute platform, combining Nvidia H200 GPUs and SambaNova SN50 RDUs, achieves 763 tokens per second on the MiniMax M2.7 mo…
Nvidia is partnering with chip startup d-Matrix to combine its GPUs with d-Matrix's Corsair inference accelerators in a joint system for AI workloads, with cloud firm Parasail as the first customer. The deal reflects Nvi…
Broadcom's stock surged following a report that Apple is developing its own AI server chip with Broadcom, while Nvidia's stock fell on the news. The deal positions Broadcom as a key supplier for Apple's AI infrastructure…
Amazon is developing a new Alexa project codenamed 'Moonraker' to give the voice assistant more advanced agentic capabilities, including handling sophisticated multistep commands. The initiative could cost over $100 mill…
Google Cloud announced the general availability of C4N, its first network- and block-storage-optimized Compute Engine instance, delivering up to 400 Gbps network bandwidth and 95 million packets per second. Built on Tita…
Meta is investing $10 billion to $13 billion in its first Canadian data center, an AI-optimized facility in Alberta powered by a new 932 MW natural gas plant, as part of a broader North American infrastructure expansion.
SK Hynix will debut on the Nasdaq on Friday with a $28 billion IPO, the largest foreign listing in history, as the memory chip maker seeks to expand capacity amid a global shortage driven by AI demand. The offering gives…
Meta Platforms is reportedly planning a large AI data center in Sturgeon County, Alberta, with an expected investment of over $10 billion and a 932 MW natural-gas-linked power component, according to Juno News and other …
Meta is building its first Canadian data center in Sturgeon County, Alberta, with a $13 billion investment. The 1GW facility, optimized for AI workloads, will create 3,000 construction jobs and 300 operational jobs, and …
XAI released Grok 4.5, a large language model trained on tens of thousands of Nvidia GB300 GPUs. Despite trailing Fable 5 and GPT-5.5 in coding benchmarks, Grok 4.5 costs $2 per million input tokens, a fraction of compet…
Nvidia's stock has lost roughly $1 trillion in market value in less than two months, now trading at 18 times forward earnings, its cheapest since early 2019. Analysts like Jessica Inskip and Vivek Arya recommend buying, …
Samsung has begun mass production of the PM1763, its first PCIe Gen6 enterprise SSD, offering sequential reads up to 28,400 MB/s and writes up to 21,000 MB/s. The drive, built with 9th-generation V-NAND and a 4nm control…