Metraipolis › archive › dc_2796a67be31a

Warpline: Binary-Level Tracing and Unified Evidence for Agentic GPU Kernel Optimization — arXiv:2611.04331

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04331
License (as archived)
CC0-1.0
Archived
2026-11-04T12:35:45.896Z
SHA-256 of the archived text
bcfad41a75c7fe49c05f7c169e1afc1b1605d247583a368912cb48951e4dec32
Stable URL of this page
https://metraipolis.com/documento/dc_2796a67be31a

How to cite

Warpline: Binary-Level Tracing and Unified Evidence for Agentic GPU Kernel Optimization — arXiv:2611.04331 — arXiv (en), https://arxiv.org/abs/2611.04331 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2796a67be31a (Metraipolis archive dc_2796a67be31a, sha256 bcfad41a75c7fe49c05f7c169e1afc1b1605d247583a368912cb48951e4dec32)
BibTeX
@misc{dc2796a67be31a,
  title = {Warpline: Binary-Level Tracing and Unified Evidence for Agentic GPU Kernel Optimization — arXiv:2611.04331},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04331},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2796a67be31a; sha256 bcfad41a75c7fe49c05f7c169e1afc1b1605d247583a368912cb48951e4dec32},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04331. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

LLM agents are becoming GPU kernel engineers, but they still tune kernels without seeing where the time goes. Hardware profilers report how busy each unit was, not where each warp's time went; intra-kernel tracers can show it, but only with hand-placed probes and a recompilation that alters the very kernel being measured. We present Warpline, which gives agents this view automatically while leaving the kernel essentially unchanged. Warpline analyzes the binary to decide which events to record and inserts probes into the binary itself without recompilation or any manual step. It then aligns the trace with hardware profiling data on one unified timeline. Because it needs only the binary, it can regenerate this evidence for every new version of the kernel at whatever level the agent edits it (source, IR, or PTX), closing the optimization loop without human intervention. On NVIDIA B200, across 35 kernels in Triton, CuTe DSL, and CUDA C++, Warpline adds a median overhead of 1.2% and leaves a median 99.9% of executed instructions unchanged, whereas existing tracers have median overheads of 1.6% to 13.2% and leave only 4.6% to 32.7% unchanged. Warpline also traces closed-source kernels, which these tracers cannot. On 10 SOL-ExecBench problems, agents using Warpline spend less time and fewer model requests than a bare coding agent and NVIDIA's Kernel Design Agents (KDA), yet lead both after every round in geometric mean and reach a 7.69$\times$ geometric-mean speedup, against 6.94$\times$ and 6.13$\times$. Two case studies further demonstrate that Warpline reveals hidden bottlenecks and provides actionable optimization insights.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.PF, nuovi lavori del giorno).
Categorie arXiv: Performance (cs.PF) ; Programming Languages (cs.PL).
Pagina del lavoro: https://arxiv.org/abs/2611.04331 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-04T12:27:47.083Z dalla pagina https://arxiv.org/list/cs.PF/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).