多智能体框架自动优化GPU核函数，性能领先多项基准测试。
KernelArc: A Multi-Agent Framework for GPU Kernel Optimization
- 多个专用智能体并行协作，仅通过结论共享和状态同步优化
- 在H100/B200上实现多项前沿算子优化，全任务榜首
- 适合需要高性能计算的算法工程师与架构研究者
我们提出KernelArc，一个用于跨异构工作负载自主优化GPU核函数的多智能体框架。策略专用智能体并行运行，通过仅传递结论的共享内存、确定性基准防护以及只读跨智能体状态与平台触发重写机制进行协调。我们在NVIDIA H100和B200 GPU上，使用代表性SOL-ExecBench工作负载评估KernelArc。生成的实现涵盖自定义BF16 GEMM、静态cuBLASLt专家接口配置表、融合专家模型反向传播、形状门控解码器层融合、原生NVFP4分组查询注意力及页式预填充注意力。在2026年8月20日发布的SOL-ExecBench公开排行榜快照中，KernelArc在所有代表性的L1、L2、量化与FlashInfer任务上均排名第一。轨迹分析支持论文核心观点：共享多智能体搜索可扩大探索范围，在固定候选预算内达到更强最优解；而各协调机制的价值取决于具体核函数与优化阶段。
原文摘要 · Abstract (English)
We present KernelArc, a multi-agent framework for autonomous GPU kernel optimization across heterogeneous workloads. Strategy-specialized agents run in parallel and coordinate through conclusions-only shared memory, a deterministic benchmark guard, and read-only cross-agent state with plateau-triggered drafting. We evaluate KernelArc on NVIDIA H100 and B200 GPUs using category-representative SOL-ExecBench workloads. The resulting implementations span custom BF16 GEMM, static cuBLASLt Expert-API configuration tables, fused mixture-of-experts backward, shape-gated decoder-layer fusion, native NVFP4 grouped-query attention, and paged prefill attention. In the public SOL-ExecBench leaderboard snapshot recorded on August~20, 2026, KernelArc ranked first on every representative L1, L2, Quantization, and FlashInfer task evaluated. The trajectories support the paper's central motivation: shared multi-agent search can broaden exploration and reach stronger incumbents within a fixed candidate budget, while the value of individual coordination features depends on the kernel and optimization stage.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。