Kernel-Smith

Kernel-Smith is a GPU kernel generation system developed by the Shanghai Artificial Intelligence Laboratory and MetaX. The technical report is available here.

We do not currently plan to release the Kernel-Smith model weights or agent code. For now, this repository will focus on sharing generated kernels, benchmarks, and related documentation. Stay tuned.

🌟 Highlights

• Uses an evolution-based optimization loop with stable evaluation on both NVIDIA Triton and MetaX MACA backends.
• Trains for kernel improvement by rewarding correctness-preserving changes that increase performance.
• Outperforms frontier models like Gemini-3.0-pro and Claude-4.6-opus on KernelBench.

🚀 Production-Level Deployment

Kernel-Smith generated kernels have already been integrated into several open-source projects:

Project	Optimized Kernel	Impact	Pull Request
SGLang	normal_decode_set_metadata	4.78x kernel acceleration	#20778
LMDeploy	DeepSeek MoE Routing	1.36x kernel acceleration	#4345
DLBlas	DeepSeek Engram kernels	Accelerated architecture research	#102

🌐 Try It Out

Kernel-Smith Online Demo