ai accelerators

Microsoft is reportedly preparing Maia 300 for a September unveiling and is seeking TSMC capacity for more than 300,000 chips to be delivered in 2027, a scale that would turn its third-generation accelerator from a limited in-house experiment into a major Azure supply program. The Information...

TrendForce says NVIDIA is evaluating lower-memory versions of its 2027 Rubin Ultra AI accelerator because the HBM supply crunch is expected to persist through that year. The immediate consequence is not a confirmed product downgrade: NVIDIA has not publicly announced a final Rubin Ultra memory...

South Korea’s National AI Computing Center has broken ground at the Solasido Data Center Park in Haenam, putting a long-delayed 15,000-accelerator public-private project on the construction path. The immediate payoff is practical: the Samsung SDS-led operator now has a route to deliver shared AI...

Business Insider’s report that AI coding agents are beginning to rewrite the software behind Nvidia’s CUDA platform points to a real change in the AI-chip market: getting a new accelerator to run an AI model is becoming faster. It does not show that CUDA has been recreated, replaced, or made...

Microwire’s argument that Nvidia’s CUDA position is being dismantled by custom AI chips, Chinese open models and domestic lithography points to real shifts in the market. But its central conclusion is ahead of the evidence. CUDA is losing exclusivity in selected workloads, not losing relevance...

Huawei’s openPangu-2.0-Pro has arrived as a 505-billion-parameter mixture-of-experts model whose technical report says full pretraining ran on Ascend 910B accelerators rather than Nvidia GPUs. The release, reported by TechTimes on August 1 and now appearing in Huawei’s Ascend Tribe community...

Samsung is reportedly developing a dedicated AI accelerator for PCs, but it is not yet a new processor family poised to displace Intel, AMD, Qualcomm, or NVIDIA in Windows machines. The chip, codenamed GAIA, is described as a standalone neural processing unit (NPU) designed to handle local AI...

Intel has unveiled Starfire, a space-grade processor that repackages its Panther Lake-era technology for U.S. government spacecraft and other systems operating beyond conventional computing environments. Detailed in an Intel product brief and reported July 13 by Tom’s Hardware and Notebookcheck...

A Chinese AI chip start-up led by semiconductor veteran Wei Shaojun publicly emerged from stealth in early July 2026, saying it will use software-defined chip design and 3D stacked near-memory computing to build domestic AI accelerators despite U.S. export controls. The South China Morning Post...

Anthropic is exploring a custom AI accelerator with Samsung Electronics, reportedly targeting Samsung’s 2-nanometer foundry process after a July 2 report said the Claude maker had entered early planning talks over chip functions, performance goals, packaging, and future server deployment. The...

Nvidia’s dominance over AMD in AI accelerators and high-end GPU computing rests less on raw silicon performance than on CUDA, the developer platform Nvidia introduced in 2006 and has spent two decades turning into the default software layer for accelerated computing. That is the uncomfortable...

President Trump said on June 18, 2026, that Apple had agreed to work with Intel to design and build chips in the United States, while the same week brought Amkor’s 10-year Arizona packaging pact with TSMC and reports of new AI-chip foundry and merchant-silicon deals. The chip industry’s week was...