<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>KV-Cache on Xu'Blog</title><link>https://xuquant.com/tags/kv-cache/</link><description>Recent content in KV-Cache on Xu'Blog</description><image><title>Xu'Blog</title><url>https://xuquant.com/og-default.png</url><link>https://xuquant.com/og-default.png</link></image><generator>Hugo -- 0.152.2</generator><language>zh</language><lastBuildDate>Sat, 11 Apr 2026 09:00:00 +0800</lastBuildDate><atom:link href="https://xuquant.com/tags/kv-cache/index.xml" rel="self" type="application/rss+xml"/><item><title>旋转约束下的压缩：从 RoPE 到 DeepSeek MLA</title><link>https://xuquant.com/posts/mathematics/position-encoding/mla-from-rope/</link><pubDate>Sat, 11 Apr 2026 09:00:00 +0800</pubDate><guid>https://xuquant.com/posts/mathematics/position-encoding/mla-from-rope/</guid><description>RoPE 与低秩压缩的不兼容性是 MLA 设计的核心驱动力——从旋转矩阵破坏低秩结构的数学证明，到解耦 RoPE 设计的工程解法。</description></item><item><title>X-Cache：小鹏自动驾驶世界模型的推理加速 Infra</title><link>https://xuquant.com/posts/world-models/xpeng-x-cache-world-model-inference-acceleration/</link><pubDate>Sat, 28 Mar 2026 18:00:00 +0800</pubDate><guid>https://xuquant.com/posts/world-models/xpeng-x-cache-world-model-inference-acceleration/</guid><description>深度解读小鹏 X-Cache：通过跨段残差缓存实现世界模型 2.7 倍推理加速，71% DiT block 跳过率且几乎零画质损失，training-free 的自动驾驶推理优化方案。</description></item><item><title>Multi-Head Latent Attention: DeepSeek V2/V3 工程视角</title><link>https://xuquant.com/posts/foundation-models/deepseek_series1_mla/</link><pubDate>Sat, 13 Sep 2025 10:00:00 +0800</pubDate><guid>https://xuquant.com/posts/foundation-models/deepseek_series1_mla/</guid><description>从 DeepSeek V2/V3 的实际部署视角分析 MLA：KV cache 压缩比、推理 throughput、与 GQA/MQA 的工程对比、长 context 下的真实收益。MLA 的数学推导见配套文章。</description></item></channel></rss>