<?xml version="1.0" encoding="utf-8"?><?xml-stylesheet type="text/xsl" href="atom.xsl"?>
<feed xmlns="http://www.w3.org/2005/Atom">
    <id>https://developer.deepx.ai/tech-docs/zh-CN/blog</id>
    <title>DEEPX Docs Blog</title>
    <updated>2026-06-18T00:00:00.000Z</updated>
    <generator>https://github.com/jpmonette/feed</generator>
    <link rel="alternate" href="https://developer.deepx.ai/tech-docs/zh-CN/blog"/>
    <subtitle>DEEPX Docs Blog</subtitle>
    <icon>https://developer.deepx.ai/tech-docs/zh-CN/img/favicon.png</icon>
    <entry>
        <title type="html"><![CDATA[Introducing DEEPX Docs]]></title>
        <id>https://developer.deepx.ai/tech-docs/zh-CN/blog/site-is-now-open</id>
        <link href="https://developer.deepx.ai/tech-docs/zh-CN/blog/site-is-now-open"/>
        <updated>2026-06-18T00:00:00.000Z</updated>
        <summary type="html"><![CDATA[We are happy to introduce DEEPX Docs, our central documentation site for DEEPX Products.]]></summary>
        <content type="html"><![CDATA[<p>We are happy to introduce DEEPX Docs, our central documentation site for DEEPX Products.</p>
<!-- -->
<p>We built this site to make technical information easier to find and maintain.</p>
<p>Here is what you can expect:</p>
<ul>
<li class="">Clear product guides and step-by-step setup instructions</li>
<li class="">Official release notes and compatibility updates</li>
<li class="">Practical troubleshooting notes and technical announcements</li>
</ul>
<p>Our goal is simple: help you spend less time searching for information and more time building with DEEPX products.</p>
<p>Thank you for visiting. More updates are coming soon.</p>]]></content>
        <author>
            <name>DEEPX Technical Writer</name>
        </author>
        <category label="Announcement" term="Announcement"/>
    </entry>
    <entry>
        <title type="html"><![CDATA[DXNN SDK v2.3.0 Released]]></title>
        <id>https://developer.deepx.ai/tech-docs/zh-CN/blog/dxnn-v2-3-0-release</id>
        <link href="https://developer.deepx.ai/tech-docs/zh-CN/blog/dxnn-v2-3-0-release"/>
        <updated>2026-04-10T00:00:00.000Z</updated>
        <summary type="html"><![CDATA[We are excited to announce DXNN SDK v2.3.0, a major release focused on runtime efficiency, security hardening, and unified application architecture. This release features significant optimizations in the YOLO post-processing pipeline and a complete architectural refresh of the DX-Stream module.]]></summary>
        <content type="html"><![CDATA[<p>We are excited to announce DXNN SDK v2.3.0, a major release focused on runtime efficiency, security hardening, and unified application architecture. This release features significant optimizations in the YOLO post-processing pipeline and a complete architectural refresh of the DX-Stream module.</p>
<!-- -->
<h2 class="anchor anchorTargetStickyNavbar_Vzrq" id="highlights">Highlights<a href="https://developer.deepx.ai/tech-docs/zh-CN/blog/dxnn-v2-3-0-release#highlights" class="hash-link" aria-label="Highlights 的直接链接" title="Highlights 的直接链接" translate="no">​</a></h2>
<ul>
<li class=""><strong>Faster YOLO post-processing</strong>: TopK-first filtering in DX-COM v2.3.0 reduces CPU overhead for DFL-based YOLO pipelines.</li>
<li class=""><strong>DX-Stream v3.0.0 refresh</strong>: Modular architecture with zero-copy (DMA-Buffer) support and no-root installation.</li>
<li class=""><strong>Unified DX-APP v3.1.0</strong>: 5-layer structure with C++/Python parity for consistent app development.</li>
<li class=""><strong>Model and tool expansion</strong>: 280+ models, 560+ examples, and new tools like <code>dxtop</code>, <code>DxScale</code>, and <code>DxConvert</code>.</li>
</ul>
<h2 class="anchor anchorTargetStickyNavbar_Vzrq" id="component-versions">Component Versions<a href="https://developer.deepx.ai/tech-docs/zh-CN/blog/dxnn-v2-3-0-release#component-versions" class="hash-link" aria-label="Component Versions 的直接链接" title="Component Versions 的直接链接" translate="no">​</a></h2>
<ul>
<li class="">DX-Compiler v2.3.0 (DX-COM v2.3.0, DX-TRON v2.0.1)</li>
<li class="">DX-Runtime v2.3.0 (DX_FW v2.5.6, NPU Driver v2.4.1, DX-RT v3.3.0, DX-Stream v3.0.0, DX-APP v3.1.0)</li>
</ul>
<h2 class="anchor anchorTargetStickyNavbar_Vzrq" id="migration-notes">Migration Notes<a href="https://developer.deepx.ai/tech-docs/zh-CN/blog/dxnn-v2-3-0-release#migration-notes" class="hash-link" aria-label="Migration Notes 的直接链接" title="Migration Notes 的直接链接" translate="no">​</a></h2>
<ul>
<li class=""><strong>DX-Stream v3.0.0</strong>: Move env vars from <code>/etc/profile.d/</code> to <code>~/.bashrc</code> and use <code>dx-modelzoo</code> for model management.</li>
<li class=""><strong>DX-APP v3.1.0</strong>: Update project layout to the new 5-layer structure.</li>
<li class=""><strong>Minimum dependency</strong>: DX-RT v3.3.0 requires NPU Driver v2.4.0+ and Firmware v2.5.2+.</li>
</ul>
<h2 class="anchor anchorTargetStickyNavbar_Vzrq" id="known-issues">Known Issues<a href="https://developer.deepx.ai/tech-docs/zh-CN/blog/dxnn-v2-3-0-release#known-issues" class="hash-link" aria-label="Known Issues 的直接链接" title="Known Issues 的直接链接" translate="no">​</a></h2>
<ul>
<li class="">PReLU-based models may still show FPS degradation.</li>
<li class="">Some models (OSNet, RepVGGA2, YoloV9C) may show accuracy variability by host CPU and calibration dataset.</li>
</ul>
<p>For full details, see <a class="" href="https://developer.deepx.ai/tech-docs/zh-CN/next/DX-AllSuite/Release_Notes">DX-AllSuite Release Notes</a>.</p>]]></content>
        <author>
            <name>DEEPX Technical Writer</name>
        </author>
        <category label="Release" term="Release"/>
    </entry>
</feed>