AISass★ 221⑂ 20 forksApache-2.0
lucialuka/OpenAssembler
GPU ML Compiler
Open source compiler translating ML repositories to SASS and CUDA binaries, achieving 3-10% speedup over Triton for LLM training and inference.
// KEY FEATURES
- Compiles ML models directly to NVIDIA SASS and CUDA binaries
- Outperforms Triton by 3-10% on LLM training and inference
- Open-source alternative to closed NVIDIA compiler toolchain
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram