{"slug": "evolving-infra-agent-adaptive-llm-serving-at-scale", "title": "Evolving Infra: Agent-Adaptive LLM Serving at Scale", "summary": "A paper posted to arXiv on 9 November 2026 (revised 12 November 2026) by Hao Kang and co-authors presents Evolving Infra, an agent-adaptive LLM serving framework that decouples compute allocation from KV cache management so nodes can switch between prefill and decode at runtime. Tested on coding, math, and computer-use tasks with GLM-5.2 and Kimi K2.6 across up to 256 GB300 GPUs, Evolving Infra cut mean end-to-end request latency by up to 4.1x and P95 time-to-first-token by up to 14x versus Dynamo, NVIDIA's production framework. The system implementation is open-sourced at github.com/mit-han-lab/EvolvingInfra.", "body_md": "# Computer Science > Distributed, Parallel, and Cluster Computing\n\n  [Submitted on 9 Nov 2026 (\n\n[v1](https://arxiv.org/abs/2611.12304v1)), last revised 12 Nov 2026 (this version, v2)]\n# Title:Evolving Infra: Agent-Adaptive LLM Serving at Scale\n\n[View PDF](https://arxiv.org/pdf/2611.12304)\n\n[HTML (experimental)](https://arxiv.org/html/2611.12304v2)\n\nAbstract:Agentic workloads are a major source of LLM inference demand. Existing prefill--decode (PD) disaggregated serving systems offer limited flexibility for runtime adaptation of compute allocation and KV cache management, and rely on coarse signals that only partly capture the diverse and dynamic agent workload. These limitations lead to load imbalance between prefill and decode nodes and eviction of KV caches that agents will reuse. We present Evolving Infra, an agent-adaptive serving framework that adapts compute allocation to agent workloads and KV cache management to agent context management at runtime. We decouple compute allocation from key--value (KV) cache management, allowing nodes to switch efficiently between prefill and decode. Building on this design, an online planner matches the PD split to workload composition, while an agent-aware distributed KV manager guides cache eviction and transfer with signals from agent context management. Across coding, math, and computer-use tasks with GLM-5.2 and Kimi K2.6 on up to 256 GB300 GPUs, Evolving Infra reduces mean end-to-end request latency and P95 TTFT by up to 4.1x and 14x over Dynamo, NVIDIA's production framework. We open-source the system implementation of Evolving Infra at [this https URL](https://github.com/mit-han-lab/EvolvingInfra).\n    \n\n## Submission history\n\nFrom: Hao Kang [\n[view email](https://arxiv.org/show-email/10ab0e7e/2611.12304)]\n\n**Mon, 9 Nov 2026 12:26:02 UTC (876 KB)**\n\n[\\[v1\\]](https://arxiv.org/abs/2611.12304v1)\n**[v2]** Thu, 12 Nov 2026 08:20:30 UTC (867 KB)\n\n### Additional Features\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/evolving-infra-agent-adaptive-llm-serving-at-scale", "canonical_source": "https://arxiv.org/abs/2611.12304", "published_at": "2026-11-16 19:43:01+00:00", "updated_at": "2026-11-16 19:54:02.669175+00:00", "lang": "en", "topics": ["ai-infrastructure", "large-language-models", "ai-agents", "ai-research", "mlops"], "entities": ["Evolving Infra", "Hao Kang", "arXiv", "GLM-5.2", "Kimi K2.6", "NVIDIA", "Dynamo", "MIT HAN Lab"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/evolving-infra-agent-adaptive-llm-serving-at-scale", "markdown": "https://wpnews.pro/news/evolving-infra-agent-adaptive-llm-serving-at-scale.md", "text": "https://wpnews.pro/news/evolving-infra-agent-adaptive-llm-serving-at-scale.txt", "jsonld": "https://wpnews.pro/news/evolving-infra-agent-adaptive-llm-serving-at-scale.jsonld"}}