Hardware
Chips, accelerators and the physical substrate AI runs on.
-
NVIDIA hands its Vera Rubin rack designs to the Open Compute Project
NVIDIA has contributed its next-generation rack and compute tray designs to the Open Compute Project, letting customers and rivals build to the…
-
Microsoft puts its own Maia silicon at the centre of inference from 2027
Microsoft has published specifications for Azure Maia 300 and said its in-house accelerators, rather than merchant GPUs, will carry the bulk of…
-
Huawei and Qualcomm sign a broad patent cross-licence covering AI
The two companies have cross-licensed their portfolios across 5G, compute, AI and networking, and Qualcomm is buying some of Huawei’s US patents,…
-
TSMC’s September revenue up 54.6 per cent on a year earlier
September revenue of NT$511.86bn was 54.6 per cent above a year earlier and fractionally below August, which is what demand running near…
-
Boston Dynamics drops a finger and nearly doubles the joints in Atlas’s hand
Atlas’s redesigned hand has four fingers and 13 degrees of freedom, and the design choices point at simulation training rather than demo…
-
TSMC September revenue up 54.6% on the year, and slightly down on August
TSMC reported September net revenue of NT$511.86bn, up 54.6% year on year but 0.6% below August, which is not the record month…
-
Samsung guides to a record quarter on AI memory demand
Samsung’s Q3 guidance puts sales near 195 trillion won and operating profit near 107.4 trillion won, against 12.17 trillion won of operating…
-
Nvidia and Microsoft put agent-scale compute on the Windows desktop
RTX Spark laptops go on sale on 16 October and a Windows version of Nvidia DGX Station enters preview, both aimed at…
-
SK hynix starts mass-producing HBM3 for NVIDIA
SK hynix began mass production of HBM3 memory, the first in the industry, and said it would supply it for NVIDIA’s H100…
-
Fugaku tops the TOP500 supercomputer list
RIKEN and Fujitsu’s Fugaku ranked first on the TOP500 list at 415.53 petaflops, and also came first on the HPCG, HPL-AI and…