{"slug": "compiler-grounded-hierarchical-diagnosis-for-llm-triton-kernel-optimization", "title": "Compiler-Grounded Hierarchical Diagnosis for LLM Triton Kernel Optimization", "summary": "Researchers have developed a compiler-grounded hierarchical optimization framework for Triton kernels that achieves a geometric-mean speedup of 4.35× and a median speedup of 2.73× across 37 benchmark entries on Ascend NPUs. The system, implemented on Triton for Ascend NPUs and evaluated on a standardized NPUKernelBench-derived Ascend 950 benchmark, escalates from pattern triage to compiler-grounded analysis to propose evidence-backed source-level rewrites. The work addresses the limitation of existing LLM-based kernel optimization approaches that rely on surface signals, which fail to explain why backend compilers miss profitable optimizations on emerging accelerators.", "body_md": "# Computer Science > Artificial Intelligence\n\n[Submitted on 25 Jul 2026]\n\n# Title:Compiler-Grounded Hierarchical Diagnosis for LLM-Based Triton Kernel Optimization\n\n[View PDF](/pdf/2607.23089)\n\n[HTML (experimental)](https://arxiv.org/html/2607.23089v1)\n\nAbstract:Recent advances in large language models (LLMs) have enabled automated kernel generation and optimization, but most existing approaches rely on surface signals such as compilation feedback and profiling metrics. These signals reveal that a kernel is slow, but not why the backend compiler fails to realize a profitable optimization, especially on emerging accelerators such as NPUs. We therefore formulate kernel optimization as a progressive cross-layer diagnosis problem that links runtime symptoms to IR structure and compiler behavior before rewriting source. Based on this insight, we present our system, a compiler-grounded and hierarchical optimization framework for Triton kernels. the system escalates from lightweight pattern triage and profiling diagnosis to IR attribution and compiler-grounded analysis only when deeper evidence is needed, then proposes evidence-backed source-level rewrites.\n\nWe implement the system on Triton for Ascend NPUs and evaluate it on 37 successfully converted entries from a standardized NPUKernelBench-derived Ascend 950 benchmark. Across these entries, the system attains a geometric-mean speedup of 4.35$\\times$ and a median speedup of 2.73$\\times$ from the initial to optimized Triton kernel; 22/37 exceed 2$\\times$ and 13/37 exceed 5$\\times$. The complete distribution ranges from near-baseline entries to large wins, motivating transparent reporting of the current system's scope and limitations.\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer\n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers\n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps\n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations\n\n*(*[What are Smart Citations?](https://www.scite.ai/))# Code, Data and Media Associated with this Article\n\nalphaXiv\n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers\n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub\n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub\n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face\n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast\n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower\n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender\n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [ Learn more about arXivLabs](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/compiler-grounded-hierarchical-diagnosis-for-llm-triton-kernel-optimization", "canonical_source": "https://arxiv.org/abs/2607.23089", "published_at": "2026-08-03 22:53:27+00:00", "updated_at": "2026-08-03 23:22:29.973674+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "large-language-models", "ai-research", "ai-infrastructure"], "entities": ["Triton", "Ascend NPUs", "NPUKernelBench", "Ascend 950"], "alternates": {"html": "https://wpnews.pro/news/compiler-grounded-hierarchical-diagnosis-for-llm-triton-kernel-optimization", "markdown": "https://wpnews.pro/news/compiler-grounded-hierarchical-diagnosis-for-llm-triton-kernel-optimization.md", "text": "https://wpnews.pro/news/compiler-grounded-hierarchical-diagnosis-for-llm-triton-kernel-optimization.txt", "jsonld": "https://wpnews.pro/news/compiler-grounded-hierarchical-diagnosis-for-llm-triton-kernel-optimization.jsonld"}}