<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Yestino - The Signal · TMA</title><link>https://yestino.com/entities/tma-4ec051</link><description>Every event involving TMA</description><language>en</language><atom:link href="https://yestino.com/entities/tma-4ec051/feed.xml" rel="self" type="application/rss+xml"/><item><title>trunk/599f186d3ff3a0e124544ecd62ca63c0a285c32d: [GPU][AOTI] Support host TMA in C++ wrappers (#193608) (#193608)</title><link>https://yestino.com/events/trunk-599f186d3ff3a0e124544ecd62ca63c0a285c32d-gpu-aoti-supp-82f959</link><guid isPermaLink="true">https://yestino.com/events/trunk-599f186d3ff3a0e124544ecd62ca63c0a285c32d-gpu-aoti-supp-82f959</guid><pubDate>Thu, 03 Sep 2026 03:24:41 GMT</pubDate><description>Summary: Host-side TensorDescriptor values created inside a torch.library.triton_op are captured through a non-Dynamo tracing path.
Sources: PyTorch Releases</description></item></channel></rss>