<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>GPU on Jerry Yang's Blog</title><link>https://blog.yangjerry.tw/tags/gpu/</link><description>Recent content in GPU on Jerry Yang's Blog</description><generator>Hugo -- gohugo.io</generator><language>zh-TW</language><managingEditor>youremail@example.com (ChengHao Yang)</managingEditor><webMaster>youremail@example.com (ChengHao Yang)</webMaster><copyright>© 2018–2025 ChengHao Yang. 除另有標示外，本站文章皆以 [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/legalcode.zh-hant) 授權釋出</copyright><lastBuildDate>Sun, 10 May 2026 12:56:48 +0800</lastBuildDate><atom:link href="https://blog.yangjerry.tw/tags/gpu/atom.xml" rel="self" type="application/rss+xml"/><item><title>初探 Kubernetes DRA —— 以 NVIDIA DRA Driver GPU 為例</title><link>https://blog.yangjerry.tw/posts/dra-nvidia-gpu-overview/</link><pubDate>Sun, 10 May 2026 12:56:48 +0800</pubDate><author>youremail@example.com (ChengHao Yang)</author><guid>https://blog.yangjerry.tw/posts/dra-nvidia-gpu-overview/</guid><description>前陣子 Dynamic Resource Allocation (DRA) 終於在 Kubernetes v1.35 進入 GA，相信不少人想要躍躍欲試，加上 NVIDIA 也把 dra-driver-nvidia-gpu 送進到 Kubernetes SIGs 當中，原有文件把 Beta 字眼拿掉，象徵技術和標準逐漸成熟。
這次我借了 CNTUG Infra Labs 目前所有的 NVIDIA GPU，來學習如何用 DRA 優雅的分配設備和資源。</description></item></channel></rss>