cuda-core

Here are 2 public repositories matching this topic...

Several optimization methods of half-precision general matrix vector multiplication (HGEMV) using CUDA core.

gpu cuda cublas nvidia gemm gemv matrix-multiply tensor-core hgemm cuda-core hgemv

Decoding Attention is specially optimized for multi head attention (MHA) using CUDA core for the decoding stage of LLM inference.

gpu cuda inference nvidia mha multi-head-attention llm large-language-model flash-attention cuda-core decoding-attention flashinfer

Add a description, image, and links to the cuda-core topic page so that developers can more easily learn about it.

To associate your repository with the cuda-core topic, visit your repo's landing page and select "manage topics."