[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-ai-agent-teams-wrote-top-ranked-gpu-kernels":10,"sections":34},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},5679,"ai-agent-teams-wrote-top-ranked-gpu-kernels","AI Agent Teams Wrote Top-Ranked GPU Kernels","KernelArc, a multi-agent AI system, topped a GPU kernel benchmark leaderboard, suggesting parallel agent search beats single-agent optimization.","A team of AI agents just wrote GPU kernels good enough to top a public benchmark leaderboard.\n\nKernelArc is a multi-agent framework that assigns different agents distinct optimization strategies and runs them in parallel to search for faster GPU kernel code. The agents share only their conclusions with each other, not their full reasoning, and coordinate through a deterministic benchmark guard plus read-only visibility into each other's progress, drafting new attempts once results plateau. Researchers tested the system on Nvidia's H100 and B200 chips using the SOL-ExecBench benchmark suite, producing kernels for tasks including custom BF16 matrix multiplication, fused mixture-of-experts backward passes, and native NVFP4 grouped-query attention. On the SOL-ExecBench public leaderboard snapshot from July 30, 2026, KernelArc's submissions ranked first across the L1, L2, Quantization, and FlashInfer categories.\n\nHand-tuning GPU kernels is slow, specialized work that directly determines how much compute, and money, an AI workload burns. If multiple agents searching in parallel genuinely outperform a single agent chasing one optimization path, that says something about how automated code optimization should be structured generally, not just for kernels.\n\nThe paper itself is careful not to oversell it: the researchers note that individual coordination features helped by varying amounts depending on the kernel and the optimization stage, a more honest caveat than most claims of AI beating humans usually include.","[\"ai\",\"gpu-kernels\",\"multi-agent\",\"nvidia\"]","2026-08-19T04:00:00.000Z","2026-08-19T10:22:34.244Z","2026-08-19T10:22:46.167Z","published",null,[],"ai",[24,26,27,28],"gpu-kernels","multi-agent","nvidia",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2608.17071",0,{"sections":35},[36,40,44,49,54,59,64,69,74,79,84,89,94,99],{"name":37,"slug":24,"count":38,"latest_published_at":39},"AI",3293,"2026-08-20T04:00:00.000Z",{"name":41,"slug":42,"count":43,"latest_published_at":39},"Security","security",435,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",210,"2026-08-19T09:32:27.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Hardware","hardware",140,"2026-08-19T18:25:42.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":63},"Consumer Tech","consumer-tech",95,"2026-08-18T16:05:00.000Z",{"name":65,"slug":66,"count":67,"latest_published_at":68},"Science","science",90,"2026-08-19T18:41:02.000Z",{"name":70,"slug":71,"count":72,"latest_published_at":73},"Software","software",73,"2026-08-18T07:51:50.000Z",{"name":75,"slug":76,"count":77,"latest_published_at":78},"Dev Tools","dev-tools",69,"2026-08-18T04:00:00.000Z",{"name":80,"slug":81,"count":82,"latest_published_at":83},"Startups","startups",47,"2026-08-19T19:13:46.000Z",{"name":85,"slug":86,"count":87,"latest_published_at":88},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",33,"2026-08-18T22:18:13.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]