Microsoft has unveiled its latest in-house AI application-specific integrated circuit (ASIC), Maia 300, and accompanying server rack, with its compute chiplets manufactured using TSMC's...
Microsoft has released architecture details for Maia 300, its third-generation Azure AI accelerator. The design pairs 96 accelerators per rack with a two-tier network that links up...
Microsoft will scale its in-house Azure Maia AI accelerators to multi-gigawatt deployments starting in 2027, making Maia the primary AI inference system for its products and services...
Samsung Electronics' next-generation 12-layer HBM4E product, a type of high-bandwidth memory (HBM), has reportedly cleared customer quality validation from Nvidia and major hyperscale...
Biren Technology is preparing its next-generation GPU for mass production, advancing a domestic AI accelerator roadmap built around broader low-precision support and locally available...
Under Lisa Su's 12-year tenure at Advanced Micro Devices (AMD), the company has gradually moved beyond its role as Intel's challenger in the CPU market. But in the accelerator market...
Data center AI accelerator package power has climbed from about 500W in the A100 generation to roughly 3,500–4,000W for Rubin Ultra-class products expected in 2027, pushing supply...
Google is preparing to send its in-house AI accelerator, the Tensor Processing Unit (TPU), into low Earth orbit for the first time as it explores whether AI computing can eventually...
The high-end cloud AI accelerator market will see a structural shift in 2027, with total ASIC shipments surpassing GPU shipments for the first time, driven by rising output from Google,...
DeepSeek is bringing more of the software used to develop its AI models to Huawei Technologies' Ascend 950 processors, deepening cooperation as Chinese technology companies seek to...
Alibaba Group is accelerating the rollout of its next-generation Zhenwu V900 accelerator by about six months as it expands a vertically integrated computing platform spanning processors,...
South Korean AI chipmaker Rebellions has expanded its neural processing units (NPUs) across four commercial SK Telecom services, with the infrastructure now processing about 14 million...
Samsung Electronics is betting on vertically stacked zHBM to move memory beyond the physical constraints of conventional 2.5D architectures, setting a system-level target of raising...
