{"type": "article", "title": "Rethinking Bursty Workloads and KV Cache Hierarchies for Efficient LLM Serving", "publisher": "Web Pulse", "url": "https://wpnews.pro/news/rethinking-bursty-workloads-and-kv-cache-hierarchies-for-efficient-llm-serving", "original_source": "https://systems.seas.harvard.edu/seminar/2026-08-03-akira-van-de-groenendaal/", "published": "2026-07-31T22:00:00+00:00", "accessed": "2026-08-01", "id": "rethinking-bursty-workloads-and-kv-cache-hierarchies-for-efficient-llm-serving"}