2+ hour, 45+ min ago   (556+ words) The release notes for CUDA NVCC 13.4.59 list supported architectures as x86_64, arm64-sbsa, and arm64 (Windows), spanning both Linux and Windows. That’s a meaningful detail for anyone tracking the toolkit’s evolution: arm64-sbsa (Server Base System Architecture) has been NVIDIA’s Linux-side Arm server target…...

7+ hour, 45+ min ago   (521+ words) shattered.io Neurophos Nets $110M, Opens Silicon Valley Office [2026] What Neurophos Just Announced The core of the September 10 announcement is a leadership expansion paired with a new office. Neurophos named Dr. Nathan Kundtz as Chief Technology Officer, promoted Dr. Michael Shebanow…...

17+ hour, 15+ min ago   (660+ words) The pitch is simple even if the hardware isn’t: pack enough memory into a single chip package that a laptop can hold a large language model entirely in memory, skip the cloud bill, and still play games on the side....

19+ hour, 43+ min ago   (524+ words) Neither Intel nor AMD has a meaningful presence in discrete GPUs at Nvidia’s scale, and neither has built the kind of software and developer ecosystem around AI acceleration that Nvidia has spent a decade assembling. A processor that pairs Nvidia’s…...

2+ day, 22+ hour ago   (687+ words) FLIP is aimed at large electricity users, chiefly data centers, that would otherwise sit in a multi-year interconnection queue waiting for SVP to confirm it has enough capacity to serve them at full, firm power. Instead, FLIP lets those customers…...

3+ day, 2+ hour ago   (342+ words) None of this confirms Astra is smarter than every rival model on every task, and the benchmark section below shows it isn’t. What it does confirm is a real infrastructure jump: coordinating more than 100,000 GPUs on one job requires networking,…...

3+ day, 7+ hour ago   (638+ words) Apple’s GPU jumped from 6 cores on the A19 Pro to 7 cores on the A20 Pro, and the company says that translates into up to 40% faster graphics performance compared with the previous generation. A single extra GPU core sounds modest on paper, but…...

4+ day, 2+ hour ago   (561+ words) AMD is done selling Nvidia’s data center customers a single chip at a time. On September 9, 2026, the clearest evidence of that shift is sitting in Oracle’s data centers: the first pieces of a planned 50,000-GPU AI supercluster built entirely on…...

4+ day, 7+ hour ago   (595+ words) What makes the Arc B390 a story now, months after that launch, is that the review cycle has matured. Early hands-on demos always favor the vendor. Independent lab testing on retail hardware does not, and that is exactly what has been…...

4+ day, 17+ hour ago   (813+ words) NVIDIA may be preparing a second version of the GeForce RTX 5070, this time built on the larger GB203 die instead of the GB205 chip that ships in the card today. The claim comes from hardware leaker hongxing2020 and has since been picked up…...