<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Yestino - The Signal · Gate CUDA TMA</title><link>https://yestino.com/zh-CN/entities/gate-cuda-tma-3d1254</link><description>与 Gate CUDA TMA 相关的全部事件</description><language>zh-CN</language><atom:link href="https://yestino.com/zh-CN/entities/gate-cuda-tma-3d1254/feed.xml" rel="self" type="application/rss+xml"/><item><title>viable/strict/1789252151: [inductor] Gate CUDA TMA tests on CUDA devices (#196636)</title><link>https://yestino.com/zh-CN/events/viable-strict-1789252151-inductor-gate-cuda-tma-tests-on-cud-1ef13b</link><guid isPermaLink="true">https://yestino.com/zh-CN/events/viable-strict-1789252151-inductor-gate-cuda-tma-tests-on-cud-1ef13b</guid><pubDate>Sat, 12 Sep 2026 18:26:42 GMT</pubDate><description>has_triton_tma_device() reports tensor-descriptor support when any installed Triton backend supports it, including CPU and XPU.
来源：PyTorch Releases</description></item></channel></rss>