<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Yestino - The Signal · TTFT</title><link>https://yestino.com/zh-CN/entities/ttft-a5626f</link><description>与 TTFT 相关的全部事件</description><language>zh-CN</language><atom:link href="https://yestino.com/zh-CN/entities/ttft-a5626f/feed.xml" rel="self" type="application/rss+xml"/><item><title>Push-based KV transfer to improve TTFT and TPOT for disaggregated inference in vLLM</title><link>https://yestino.com/zh-CN/events/push-based-kv-transfer-to-improve-ttft-and-tpot-for-disaggre-20aa1b</link><guid isPermaLink="true">https://yestino.com/zh-CN/events/push-based-kv-transfer-to-improve-ttft-and-tpot-for-disaggre-20aa1b</guid><pubDate>Mon, 19 Oct 2026 16:05:17 GMT</pubDate><description>vLLM&apos;s prefill-decode disaggregation splits LLM inference across engine instances but must transfer the KV cache between them.
来源：AWS Machine Learning</description></item></channel></rss>