Year: 2025
-
Arm GPUs to Add Neural Networks to Graphics Processing
Arm is extending Vulkan to add neural networks to graphics processing. To be supported in its next-generation GPU, the Arm neural technology can reduce power. Arm’s first model, Neural Super Sampling (NSS), upscales images to a higher resolution. Future models will interpolate frames and improve ray-traced image quality.
-
SMT Offers a Choice Between Single-Thread and Total Throughput
Symmetric multithreading (SMT) may up total CPU throughput, but it slows each thread. It offers a valuable choice between total and single-thread performance.
-
OpenInfer Eases AI-Assistant Development
Edge AI promises the benefits of cloud AI, plus responsiveness and privacy, because local model execution eliminates sending personal information to the cloud and waiting for a reply. However, edge systems’ variety challenges developers. They differ in supported data types, processing power, memory capacity, and connected sensors. The Silicon Valley startup OpenInfer addresses the edge…
-
Premiere: Intel 2Q 2025 Earnings Call Review
I’m filing this under Site News. I shot and posted my first YouTube video. It covers the same content as the 2Q25 Intel earnings article, aiming to reach a new audience. Thanks for supporting XPU.pub over the past 1.5 years, and I appreciate any support you can lend to the new format. Video quality should…
-
Intel Backs 18A and 14A Development in Q2 2025 Call
Intel’s CEO, Lip-Bu Tan, voiced a disciplined capex strategy, approving 18A and 14A investments only after yields, performance, and customers meet targets.
-
Cuda Comes to RISC-V Hosts, Not Devices
Nvidia intends to port Cuda to RISC-V, but this is less revolutionary than it sounds. Speaking at the Fifth RISC-V China Summit, Nvidia VP Frans Sijstermans stated the company is cooperating with hardware partners to support the open architecture in the standard Cuda version. However, this doesn’t imply Nvidia will support RISC-V chips as alternatives…