<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic [Android 14/15/16] TensorFlow Lite GPU delegate not supported? (ObjectDetection demo issue) in i.MX Processors</title>
    <link>https://community.nxp.com/t5/i-MX-Processors/Android-14-15-16-TensorFlow-Lite-GPU-delegate-not-supported/m-p/2201727#M242040</link>
    <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;I have been testing the TensorFlow Lite &lt;EM&gt;Object Detection&lt;/EM&gt; demo on several Android BSP releases for the i.MX8MQ EVK, but the GPU delegate cannot be enabled successfully.&lt;BR /&gt;The inference works correctly on CPU, but when switching to GPU delegate, initialization fails and inference does not start.&lt;/P&gt;&lt;H4&gt;&lt;STRONG&gt;Test Environment&lt;/STRONG&gt;&lt;/H4&gt;&lt;DIV class=""&gt;&lt;DIV class=""&gt;Item Version / Detail &lt;TABLE&gt;&lt;TBODY&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;SoC&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;NXP i.MX8MQ&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;Android BSP&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;android-16.0.0_1.0.0 / android-15.0.0_2.0.0 / android-14.0.0_1.2.0&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;Demo&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;TensorFlow Lite Object Detection (from TensorFlow)&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;GPU&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;Vivante GC7000Lite&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;NPU&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;Not used&lt;/TD&gt;&lt;/TR&gt;&lt;/TBODY&gt;&lt;/TABLE&gt;&lt;H4&gt;&amp;nbsp;&lt;/H4&gt;&lt;H4&gt;&lt;STRONG&gt;Problem Details&lt;/STRONG&gt;&lt;/H4&gt;&lt;P&gt;The app runs fine with CPU inference, but when enabling GPU delegate, it fails to initialize.&lt;BR /&gt;Here are some of the relevant logcat messages during GPU mode:&lt;/P&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;LI-CODE lang="python"&gt;2025-09-11 20:05:58.595  libEGL  E  call to OpenGL ES API with no current context (logged once per thread)
2025-09-11 20:05:55.994  Surface  E  IGraphicBufferProducer::setBufferCount(0) returned Invalid argument
2025-09-11 20:05:56.086  BaseTaskApi  W  Closing an already closed native lib&lt;/LI-CODE&gt;&lt;P&gt;in the code:&lt;/P&gt;&lt;LI-CODE lang="cpp"&gt;                if (CompatibilityList().isDelegateSupportedOnThisDevice) {
                    baseOptionsBuilder.useGpu()
                    activeDelegate = "GPU"
                } else {
                    objectDetectorListener?.onError("GPU is not supported on this device") // also print this !!!!!!!!!!!!
                }&lt;/LI-CODE&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;I have check:&lt;/P&gt;&lt;DIV&gt;In /vendor/lib64/, there are libraries such as &lt;STRONG&gt;libOpenCL.so&lt;/STRONG&gt;, &lt;STRONG&gt;libtim-vx.so&lt;/STRONG&gt;,&lt;/DIV&gt;&lt;DIV&gt;&amp;nbsp;&lt;/DIV&gt;&lt;DIV&gt;&lt;DIV&gt;&lt;SPAN class=""&gt;&lt;SPAN class=""&gt;Is direct GPU Delegate unsupported on i.MX8MQ even in Android&amp;nbsp; BSP? (Similar to &lt;/SPAN&gt;&lt;/SPAN&gt;&lt;A href="https://community.nxp.com/t5/i-MX-Processors/i-MX8MQ-Android-13-can-t-use-GPU-Delegate-to-inference-AI-model/td-p/1908208" target="_blank" rel="noopener noreferrer nofollow"&gt;&lt;SPAN class=""&gt;this Android 13 thread&lt;/SPAN&gt;&lt;/A&gt;&lt;SPAN class=""&gt;&lt;SPAN class=""&gt;, where VX was recommended for Android 14+.)&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
    <pubDate>Mon, 10 Nov 2025 09:22:50 GMT</pubDate>
    <dc:creator>hank2024</dc:creator>
    <dc:date>2025-11-10T09:22:50Z</dc:date>
    <item>
      <title>[Android 14/15/16] TensorFlow Lite GPU delegate not supported? (ObjectDetection demo issue)</title>
      <link>https://community.nxp.com/t5/i-MX-Processors/Android-14-15-16-TensorFlow-Lite-GPU-delegate-not-supported/m-p/2201727#M242040</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;I have been testing the TensorFlow Lite &lt;EM&gt;Object Detection&lt;/EM&gt; demo on several Android BSP releases for the i.MX8MQ EVK, but the GPU delegate cannot be enabled successfully.&lt;BR /&gt;The inference works correctly on CPU, but when switching to GPU delegate, initialization fails and inference does not start.&lt;/P&gt;&lt;H4&gt;&lt;STRONG&gt;Test Environment&lt;/STRONG&gt;&lt;/H4&gt;&lt;DIV class=""&gt;&lt;DIV class=""&gt;Item Version / Detail &lt;TABLE&gt;&lt;TBODY&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;SoC&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;NXP i.MX8MQ&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;Android BSP&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;android-16.0.0_1.0.0 / android-15.0.0_2.0.0 / android-14.0.0_1.2.0&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;Demo&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;TensorFlow Lite Object Detection (from TensorFlow)&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;GPU&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;Vivante GC7000Lite&lt;/TD&gt;&lt;/TR&gt;&lt;TR&gt;&lt;TD width="104.915px" height="24px"&gt;NPU&lt;/TD&gt;&lt;TD width="501.364px" height="24px"&gt;Not used&lt;/TD&gt;&lt;/TR&gt;&lt;/TBODY&gt;&lt;/TABLE&gt;&lt;H4&gt;&amp;nbsp;&lt;/H4&gt;&lt;H4&gt;&lt;STRONG&gt;Problem Details&lt;/STRONG&gt;&lt;/H4&gt;&lt;P&gt;The app runs fine with CPU inference, but when enabling GPU delegate, it fails to initialize.&lt;BR /&gt;Here are some of the relevant logcat messages during GPU mode:&lt;/P&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;LI-CODE lang="python"&gt;2025-09-11 20:05:58.595  libEGL  E  call to OpenGL ES API with no current context (logged once per thread)
2025-09-11 20:05:55.994  Surface  E  IGraphicBufferProducer::setBufferCount(0) returned Invalid argument
2025-09-11 20:05:56.086  BaseTaskApi  W  Closing an already closed native lib&lt;/LI-CODE&gt;&lt;P&gt;in the code:&lt;/P&gt;&lt;LI-CODE lang="cpp"&gt;                if (CompatibilityList().isDelegateSupportedOnThisDevice) {
                    baseOptionsBuilder.useGpu()
                    activeDelegate = "GPU"
                } else {
                    objectDetectorListener?.onError("GPU is not supported on this device") // also print this !!!!!!!!!!!!
                }&lt;/LI-CODE&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;I have check:&lt;/P&gt;&lt;DIV&gt;In /vendor/lib64/, there are libraries such as &lt;STRONG&gt;libOpenCL.so&lt;/STRONG&gt;, &lt;STRONG&gt;libtim-vx.so&lt;/STRONG&gt;,&lt;/DIV&gt;&lt;DIV&gt;&amp;nbsp;&lt;/DIV&gt;&lt;DIV&gt;&lt;DIV&gt;&lt;SPAN class=""&gt;&lt;SPAN class=""&gt;Is direct GPU Delegate unsupported on i.MX8MQ even in Android&amp;nbsp; BSP? (Similar to &lt;/SPAN&gt;&lt;/SPAN&gt;&lt;A href="https://community.nxp.com/t5/i-MX-Processors/i-MX8MQ-Android-13-can-t-use-GPU-Delegate-to-inference-AI-model/td-p/1908208" target="_blank" rel="noopener noreferrer nofollow"&gt;&lt;SPAN class=""&gt;this Android 13 thread&lt;/SPAN&gt;&lt;/A&gt;&lt;SPAN class=""&gt;&lt;SPAN class=""&gt;, where VX was recommended for Android 14+.)&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Mon, 10 Nov 2025 09:22:50 GMT</pubDate>
      <guid>https://community.nxp.com/t5/i-MX-Processors/Android-14-15-16-TensorFlow-Lite-GPU-delegate-not-supported/m-p/2201727#M242040</guid>
      <dc:creator>hank2024</dc:creator>
      <dc:date>2025-11-10T09:22:50Z</dc:date>
    </item>
    <item>
      <title>Re: [Android 14/15/16] TensorFlow Lite GPU delegate not supported? (ObjectDetection demo issue)</title>
      <link>https://community.nxp.com/t5/i-MX-Processors/Android-14-15-16-TensorFlow-Lite-GPU-delegate-not-supported/m-p/2205820#M242207</link>
      <description>&lt;P&gt;Hi hank:&lt;/P&gt;
&lt;P&gt;The stock TFLite GPU delegate is not supported on i.MX8MQ Android BSPs , this is expected due to GPU backend compatibility and NXP's chosen acceleration strategy.&amp;nbsp; Use VX (TIM-VX) delegate on i.MX8M, and Neutron delegate on i.MX95.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Mon, 17 Nov 2025 06:14:00 GMT</pubDate>
      <guid>https://community.nxp.com/t5/i-MX-Processors/Android-14-15-16-TensorFlow-Lite-GPU-delegate-not-supported/m-p/2205820#M242207</guid>
      <dc:creator>danielchen</dc:creator>
      <dc:date>2025-11-17T06:14:00Z</dc:date>
    </item>
    <item>
      <title>Re: [Android 14/15/16] TensorFlow Lite GPU delegate not supported? (ObjectDetection demo issue)</title>
      <link>https://community.nxp.com/t5/i-MX-Processors/Android-14-15-16-TensorFlow-Lite-GPU-delegate-not-supported/m-p/2205889#M242211</link>
      <description>&lt;P&gt;Thank you very much for your reply. I have already resolved the issue. I also referred to this document.&amp;nbsp;&lt;A href="https://www.nxp.com.cn/docs/en/user-guide/IMX_ANDROID_TENSORFLOWLITE_USERS_GUIDE.pdf" target="_blank"&gt;i.MX TensorFlow Lite on Android User's Guide&lt;/A&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Mon, 17 Nov 2025 07:55:12 GMT</pubDate>
      <guid>https://community.nxp.com/t5/i-MX-Processors/Android-14-15-16-TensorFlow-Lite-GPU-delegate-not-supported/m-p/2205889#M242211</guid>
      <dc:creator>hank2024</dc:creator>
      <dc:date>2025-11-17T07:55:12Z</dc:date>
    </item>
  </channel>
</rss>

