Latest about CUDA
A project to bring CUDA to non-Nvidia GPUs is making major progress
ZLUDA has expanded its team and made steady technical progress in Q1: Expands support for LLM.c, attempts to bring back PhysX 32-bit.
"Export control was a failure." Nvidia CEO Jensen Huang slams Biden's China AI restrictions - Computex 2025 Q and A
Premium Nvidia's CEO Jensen Huang addressed trade and new AI technologies in a Computex Q&A
Nvidia to drop CUDA support for Maxwell, Pascal, and Volta GPUs with the next major Toolkit release
Nvidia indirectly prompts developers to upgrade as Maxwell (GTX 900), Pascal (GTX 1000), and Volta (data-center) GPUs will not be supported in the next major CUDA version.
China's Moore Threads polishes homegrown CUDA alternative
Moore Threads' MUSA shows China's intent to break away from CUDA with its own GPU programming environments.
Keller and Koduri headline the Beyond CUDA Summit today — AI leaders rally to challenge Nvidia's dominance
The Beyond CUDA Summit commences tomorrow, bringing together key leaders and engineers in the industry to explore the future of AI with alternative platforms.
PhysX quietly retired on RTX 50 series GPUs: Nvidia ends 32-bit CUDA app support
With the retirement of 32-bit CUDA application support on RTX 50 series GPUs, PhysX is now end-of-life starting with Blackwell and newer Nvidia GPU architectures.
DeepSeek's AI breakthrough bypasses industry-standard CUDA for some functions, uses Nvidia's assembly-like PTX programming instead
DeepSeek used PTX ISA to fine-tune its AI model training.
RTX 5090 prototype allegedly has 24,576 CUDA cores and 800W TDP
Hardware leaker HXL spotted an alleged RTX 5090 prototype on the Chinese Chiphell forums.
