how-to-optim-algorithm-in-cuda

RAFT contains fundamental widely-used algorithms and primitives for machine learning and information retrieval. The algorithms are CUDA-accelerated and form building blocks for more easily writing high performance applications.

FlashInfer: Kernel Library for LLM Serving

hpc 教程，包含集合通信(mpi、nccl)、cuda 编程、向量化 SIMD、RDMA 通信等

Mirage Persistent Kernel: Compiling LLMs into a MegaKernel