DeepGEMM
Unlock Maximum FP8 Performance on Hopper GPUs
Penerangan
DeepGEMM, from DeepSeek, is an open-source library for highly optimized FP8 GEMM kernels on Hopper GPUs. Clean codebase (~300 LOC), JIT-compiled, no heavy dependencies.
DeepGEMM, from DeepSeek, is an open-source library for highly optimized FP8 GEMM kernels on Hopper GPUs. Clean codebase (~300 LOC), JIT-compiled, no heavy dependencies.
