KernelSkill: A Multi-Agent Framework for GPU Kernel Optimization

• URL: http://arxiv.org/abs/2603.10085v1
• Date: Tue, 10 Mar 2026 13:43:38 GMT
• Title: KernelSkill: A Multi-Agent Framework for GPU Kernel Optimization
• Abstract summary: We present KernelSkill, a multi-agent framework with a dual-level memory architecture.<n>On Kernel Levels 1-3, KernelSkill achieves a 100% success rate and average speedups of 5.44xBench, 2.82x, and 1.92x over Torch Eager.
• Score: 37.008713142905336
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Improving GPU kernel efficiency is crucial for advancing AI systems. Recent work has explored leveraging large language models (LLMs) for GPU kernel generation and optimization. However, existing LLM-based kernel optimization pipelines typically rely on opaque, implicitly learned heuristics within the LLMs to determine optimization strategies. This leads to inefficient trial-and-error and weakly interpretable optimizations. Our key insight is to replace implicit heuristics with expert optimization skills that are knowledge-driven and aware of task trajectories. Specifically, we present KernelSkill, a multi-agent framework with a dual-level memory architecture. KernelSkill operates by coordinating agents with long-term memory of reusable expert skills and short-term memory to prevent repetitive backtracking. On KernelBench Levels 1-3, KernelSkill achieves a 100% success rate and average speedups of 5.44x, 2.82x, and 1.92x over Torch Eager on Levels 1, 2, and 3, respectively, outperforming prior baselines. Code is available at https://github.com/0satan0/KernelMem/.

Related papers

• HIERA: Workload-Aware Planning Across Implementation Spaces for GPU Kernel Optimization [17.600449573236194]
  We propose textscHIERA, a hierarchical search-space planning framework for kernel optimization.<n>textscHIERA constructs contract-augmented task specifications, selects an appropriate implementation space across PyTorch operators, libraries, and custom kernels.<n>Experiments show textscHIERA delivers stronger overall implementation refinement, validity, and optimization performance than existing training-free methods.
  arXiv  Detail & Related papers  (2026-08-21T14:29:20Z)
• GPU Forecasters: Language Models as Selective Surrogates for Kernel Runtime Optimization [73.92934811090163]
  We study how LLMs can serve as selective GPU surrogates for kernel evaluation.<n>We measure whether their forecasts are accurate, calibrated, and practically useful for recovering fast kernels under limited GPU-measurement budgets.<n>Our experiments demonstrate that LLMs can accurately forecast relative kernel performance, that their utility can be improved through reinforcement learning.
  arXiv  Detail & Related papers  (2026-05-29T15:56:08Z)
• Learning When to Optimize: Verified Optimization Skills from Expert GPU-Kernel Lineages [4.0702755941542925]
  LLM-based agents are increasingly used to generate GPU kernels.<n>They often know what optimizations to try without knowing when those optimizations are sound.<n>We introduce KLineage, which learns this missing "when" knowledge from expert kernels.
  arXiv  Detail & Related papers  (2026-05-27T09:32:16Z)
• Making LLMs Optimize Multi-Scenario CUDA Kernels Like Experts [25.264078143000045]
  Current automated kernel optimization methods focus on machine learning applications, while overlooking broader domains like sparse matrix operations in scientific computing.<n>In this paper, we address the absence of systematic evaluation for multi-scenario settings introducing MSBench.<n>Building on this benchmark, we introduce MSMaster, a hardware-aware system for kernel optimization that leverages GPU profiling information and automatically constructs the full compilation and execution toolchain.
  arXiv  Detail & Related papers  (2026-03-07T12:22:43Z)
• CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation [51.72529978689561]
  Agent is a large-scale agentic reinforcement learning system that develops kernel expertise through three components.<n>Agent delivers 100%, 100%, and 92% faster rate over torchcompile on KernelBench.
  arXiv  Detail & Related papers  (2026-02-27T18:58:05Z)
• KernelBlaster: Continual Cross-Task CUDA Optimization via Memory-Augmented In-Context Reinforcement Learning [3.4998382481249286]
  We release KernelBlaster as an open-source agentic framework, accompanied by a test harness, verification components, and a reproducible evaluation.<n>Our method achieves mean speed-ups of 1.43x, 2.50x, and 1.50x on KernelBench Levels 1, 2, and 3, respectively.
  arXiv  Detail & Related papers  (2026-02-15T19:48:43Z)
• Optimizing PyTorch Inference with LLM-Based Multi-Agent Systems [1.2289544895833646]
  We present a framework for comparing multi-agent PyTorch optimization systems.<n>We show that exploit-heavy strategies perform best when paired with error-fixing agents.<n>The best implementation achieves an average 2.88x speedup on an H100 GPU.
  arXiv  Detail & Related papers  (2025-11-21T05:37:38Z)
• STARK: Strategic Team of Agents for Refining Kernels [23.717055490630596]
  We introduce an agentic framework for GPU kernel optimization that explores the design space through multi-agent collaboration.<n>This framework mimics the workflow of expert engineers, enabling LLMs to reason about hardware trade-offs, incorporate profiling feedback, and refine kernels iteratively.<n>We evaluate our approach on KernelBench, a benchmark for LLM-based kernel optimization, and demonstrate substantial improvements over baseline agents.
  arXiv  Detail & Related papers  (2025-10-19T20:41:46Z)
• SwizzlePerf: Hardware-Aware LLMs for GPU Kernel Performance Optimization [4.043444100380313]
  SwizzlePerf automatically generates spatial optimizations for GPU kernels on disaggregated architectures.<n>On a suite of 10 diverse ML and Science kernels, SwizzlePerf can generate swizzling patterns for 9 of the kernels that achieve up to a 2.06x speedup and 70% improvement in L2 hit rate.
  arXiv  Detail & Related papers  (2025-08-27T20:30:43Z)
• vTensor: Flexible Virtual Tensor Management for Efficient LLM Serving [53.972175896814505]
  Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests. Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests.
  arXiv  Detail & Related papers  (2024-07-22T14:37:58Z)
• Harnessing Deep Learning and HPC Kernels via High-Level Loop and Tensor Abstractions on CPU Architectures [67.47328776279204]
  This work introduces a framework to develop efficient, portable Deep Learning and High Performance Computing kernels. We decompose the kernel development in two steps: 1) Expressing the computational core using Processing Primitives (TPPs) and 2) Expressing the logical loops around TPPs in a high-level, declarative fashion. We demonstrate the efficacy of our approach using standalone kernels and end-to-end workloads that outperform state-of-the-art implementations on diverse CPU platforms.
  arXiv  Detail & Related papers  (2023-04-25T05:04:44Z)
• Kernel methods through the roof: handling billions of points efficiently [94.31450736250918]
  Kernel methods provide an elegant and principled approach to nonparametric learning, but so far could hardly be used in large scale problems. Recent advances have shown the benefits of a number of algorithmic ideas, for example combining optimization, numerical linear algebra and random projections. Here, we push these efforts further to develop and test a solver that takes full advantage of GPU hardware.
  arXiv  Detail & Related papers  (2020-06-18T08:16:25Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.