llama.cpp b10724 发布：KV 缓存恢复性能大幅优化

AI 摘要

llama.cpp 发布 b10724 版本，优化了 KV 缓存状态恢复时非连续单元格的散列读取性能，通过按连续运行批量复制，将恢复时间从 25-63 秒降至 221-424 毫秒，并增加了相关测试。

正文节选

<details open> kv-cache : optimize restoring non-contiguous cells (#27991) * kv cache : batch state restore scatter reads per contiguous run When restoring state into non-contiguous destination cells (e.g. a prompt-cache snapshot into a fragmented ring), state_read_data issued one small copy per KV cell - ~1.4M copies of a few KiB each for a 40k+ token restore, taking 25-63 s on the CUDA backend. The snapshot stores cell rows in cell order, so a maximal run of consecutive destination indices

发布时间：2026-09-01 01:26

抓取时间：2026-09-01 02:32

来源机构：ggml-org