AutoKernel: Autonomous GPU Kernel Optimization via Iterative Agent-Driven Search

TL;DR

AutoKernel is an autonomous framework that optimizes GPU kernels for machine learning models through iterative, agent-driven search, significantly improving performance without human intervention.

Contribution

It introduces a fully automated GPU kernel optimization system that profiles, tests, and refines kernels for PyTorch models using an iterative agent loop.

Findings

AutoKernel's Triton kernels outperform PyTorch eager and torch.compile.

Achieved up to 5.29x speedup on RMSNorm.

Won first place on vectorsum_v2 B200 leaderboard.

Abstract

Writing high-performance GPU kernels is among the most labor-intensive tasks in machine learning systems engineering. We present AutoKernel, an open-source framework that applies an autonomous agent loop to GPU kernel optimization for arbitrary PyTorch models. Given a model, AutoKernel profiles it to identify computational bottlenecks, ranks them by Amdahl's law impact, and iteratively refines Triton or CUDA C++ kernel implementations through hundreds of experiments without human intervention. A five-stage correctness harness covering smoke tests, shape sweeps, numerical stability, determinism verification, and edge-case coverage ensures every candidate kernel is validated before any speedup is recorded. The system comprises over 9,000 lines of Python, 18 starter kernel implementations across two backends, a six-tier optimization playbook, and integration with the KernelBench benchmark suite. AutoKernel covers nine kernel types spanning the dominant operations in modern transformer architectures. On an NVIDIA H100, our Triton kernels outperform both PyTorch eager and torch.compile (max-autotune) on the majority of tested configurations: 5.29x over eager on RMSNorm, 2.82x on softmax, and 2.21x on cross-entropy, while beating torch.compile by 2.83x, 3.44x, and 2.94x respectively. In community deployment, an AutoKernel-optimized kernel achieved first place on the vectorsum_v2 B200 leaderboard. The full system is available at https://github.com/RightNow-AI/autokernel.