megakernel
AutoMegaKernel
An agent harness that compiles a model into one provably-correct, self-retargeting CUDA megakernel and self-tunes it past cuBLAS at batch-1 LLM decode, paper: https://arxiv.org/abs/2606.09682
- 166
- Python
- MIT
- 2 mo ago
bonsai-turbo
Single-launch batch-1 decode engine for PrismML Bonsai 27B (ternary and 1-bit) on NVIDIA GPUs. 1.76x the vendor llama.cpp fork on H100, same outputs.
- 19
- Cuda
- Apache-2.0
- 4 mo ago


