{"slug": "towards-efficient-hpc-systems-for-agents-challenges-and-opportunities", "title": "Towards Efficient HPC Systems for Agents: Challenges and Opportunities", "summary": "A September 30, 2026 arXiv paper measuring coding-agent use of high-performance computing systems found that users running coding agents were only 19.5% of the observed population yet generated 55.8% of job submissions, 29.1% of CPU core-hours, and 42.7% of GPU-hours. The authors report agents issue commands at 20.8x the human rate and decompose work into fine-grained explore-modify-execute loops, straining schedulers, bandwidth-provisioned filesystems, and creating new prompt-injection and policy-enforcement surfaces. The paper argues HPC facilities should treat agents as first-class principals and co-design compute, storage, agent memory, and safety rather than banning agents or treating them as ordinary users.", "body_md": "# Computer Science > Distributed, Parallel, and Cluster Computing\n\n  [Submitted on 30 Sep 2026]\n\n# Title:Towards Efficient HPC Systems for Agents: Challenges and Opportunities\n\n[View PDF](https://arxiv.org/pdf/2609.38723)\n\n[HTML (experimental)](https://arxiv.org/html/2609.38723v1)\n\nAbstract:Coding agents have become real users of high-performance computing (HPC) systems, yet today's HPC abstractions, interfaces, and policies remain designed for human-driven workflows. In our measurement, users running coding agents are only 19.5% of the observed population, but account for 55.8% of job submissions, 29.1% of CPU core-hours, and 42.7% of GPU-hours. Agents are not simply faster humans. They issue commands at 20.8x the human rate, decompose work into fine-grained explore-modify-execute loops, and pursue open-ended goals through trial-and-error campaigns that continue through nights and weekends. These behaviors strain abstractions built for human timescales, surfacing as control-plane pressure on the scheduler, metadata-intensive I/O on bandwidth-provisioned filesystems, repeated rediscovery of what earlier sessions already learned, and new prompt-injection and policy-enforcement surfaces. Neither banning agents nor treating them as ordinary users is sustainable. We instead argue for co-design, that facilities should treat agents as first-class principals where agents become facility-aware tenants. We outline the resulting challenges and opportunities in compute, storage, agent memory, and safety.\n    \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/))\n# 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))\n# 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))\n# 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/towards-efficient-hpc-systems-for-agents-challenges-and-opportunities", "canonical_source": "https://arxiv.org/abs/2609.38723", "published_at": "2026-10-08 03:48:21+00:00", "updated_at": "2026-10-08 04:20:33.852966+00:00", "lang": "en", "topics": ["ai-agents", "ai-infrastructure", "ai-safety", "ai-research"], "entities": ["arXiv", "HPC systems", "coding agents"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/towards-efficient-hpc-systems-for-agents-challenges-and-opportunities", "markdown": "https://wpnews.pro/news/towards-efficient-hpc-systems-for-agents-challenges-and-opportunities.md", "text": "https://wpnews.pro/news/towards-efficient-hpc-systems-for-agents-challenges-and-opportunities.txt", "jsonld": "https://wpnews.pro/news/towards-efficient-hpc-systems-for-agents-challenges-and-opportunities.jsonld"}}