Flat skin for active admin.
Repository Index
47,133 repositories indexed
1–2 of 2 · Most stars
Open source Python/CUDA C/C++ to SASS/CUDA Binary compiler (previously closed source by nvidia), ~16.5% faster than PyTorch CUDA API in ML tasks, tie PTXAS on ReLU, GeLU, Adam etc. Open for specific kernel optimization that break the existing categories further down into cases as well as architectural innovations.
- compiler
- compiler-backend
- compiler-optimization
- gpu-acceleration