An evolutionary framework for GPU kernel generation that outperforms frontier models like Claude 4.6 and Gemini 3.0.
March 31, 2026
Original Paper
Kernel-Smith: A Unified Recipe for Evolutionary Kernel Optimization
arXiv · 2603.28342
The Takeaway
Kernel-Smith uses an evolution-in-the-loop agent to generate high-performance Triton code, beating the best proprietary LLMs at low-level operator optimization. This is a significant step toward automated, cross-platform hardware acceleration and high-utilization ML infra.
From the abstract
We present Kernel-Smith, a framework for high-performance GPU kernel and operator generation that combines a stable evaluation-driven evolutionary agent with an evolution-oriented post-training recipe. On the agent side, Kernel-Smith maintains a population of executable candidates and iteratively improves them using an archive of top-performing and diverse programs together with structured execution feedback on compilation, correctness, and speedup. To make this search reliable, we build backend