CUDA
Tag804 stories
CUDA news and updates covering NVIDIA's platform for general-purpose computing on GPUs. Readers can learn about the programming model and memory hierarchy, kernel optimisation and occupancy, libraries such as cuBLAS and cuDNN, toolkit and driver compatibility, and portability to other accelerators.
ZLUDA Has Been Seeing New Activity For CUDA On AMD GPUsTiny LLM hacks: Loading quantized model using Python/llama_cpp_python.Announcing Confidential Computing General Access on NVIDIA H100 Tensor Core GPUsThe One Billion Row Challenge in CUDA: from 17m to 17sllm.ckarpathy/llm.c: LLM training in simple, raw C/CUDAAMD Releases Orochi 2.0 With More CUDA/HIP Functions Implemented For Better PortabilityNVIDIA Wants More Programming Languages to Support CUDAEfficient CUDA Debugging: Using NVIDIA Compute Sanitizer with NVIDIA Tools Extension and Creating Custom Toolsmlecauchois/micrograd-cuda
