<?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>Triton on 兜忆轩 | Cedric's Compiler Notes</title><link>https://douyixuan.github.io/tags/triton/</link><description>Recent content in Triton on 兜忆轩 | Cedric's Compiler Notes</description><image><title>兜忆轩 | Cedric's Compiler Notes</title><url>https://douyixuan.github.io/images/papermod-cover.png</url><link>https://douyixuan.github.io/images/papermod-cover.png</link></image><generator>Hugo -- 0.146.0</generator><language>en-us</language><lastBuildDate>Fri, 11 Sep 2026 18:41:00 +0800</lastBuildDate><atom:link href="https://douyixuan.github.io/tags/triton/index.xml" rel="self" type="application/rss+xml"/><item><title>CUDA 软件栈测试方法论：从 Differential Testing 到 GPU Compiler Fuzzing</title><link>https://douyixuan.github.io/posts/compiler/cuda-software-stack-testing-and-fuzzing/</link><pubDate>Fri, 11 Sep 2026 18:41:00 +0800</pubDate><guid>https://douyixuan.github.io/posts/compiler/cuda-software-stack-testing-and-fuzzing/</guid><description>从 CUDA 软件栈测试全貌出发，给出一套面向 Triton/GPU compiler 的 fuzz 测试选型、架构、CI 分层和落地方案。</description></item><item><title>CUDA 软件栈测试方法论：从 Differential Testing 到 GPU Compiler Fuzzing</title><link>https://douyixuan.github.io/posts/compiler/cuda-stack-testing-and-fuzzing/</link><pubDate>Fri, 11 Sep 2026 18:40:00 +0800</pubDate><guid>https://douyixuan.github.io/posts/compiler/cuda-stack-testing-and-fuzzing/</guid><description>从 CUDA 软件栈测试全貌出发，给出一套面向 Triton/GPU Compiler 的 fuzz 选型与可落地方案：Hypothesis 生成合法程序，Reference/Differential 作为 oracle，Compute Sanitizer 检查内存与并发，mlir-reduce 自动缩减失败用例。</description></item></channel></rss>