<?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 Does eIQ support Google Edge TPU (through Accelerator Module or USB Accelerator)? in i.MX RT Crossover MCUs</title>
    <link>https://community.nxp.com/t5/i-MX-RT-Crossover-MCUs/Does-eIQ-support-Google-Edge-TPU-through-Accelerator-Module-or/m-p/1489590#M20613</link>
    <description>&lt;P&gt;I already tried to ask &lt;A href="https://stackoverflow.com/questions/71590845/does-tensorflow-lite-for-microcontrollers-support-google-edge-tpu/71685429" target="_self"&gt;here&lt;/A&gt;, but didn't get a satisfying answer. I'll try again in here since I want to use the i.MX RT1170.&lt;/P&gt;&lt;P&gt;I already know that TensorFlow Lite (TFL)&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://coral.ai/docs/edgetpu/models-intro/" target="_blank" rel="nofollow noopener noreferrer"&gt;supports&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;the Google Edge TPU, for instance through the&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://coral.ai/products/dev-board/" target="_blank" rel="nofollow noopener noreferrer"&gt;Coral Dev Board&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;(Linux required).&lt;/P&gt;&lt;P&gt;However I'd like to know whether&lt;SPAN&gt;&amp;nbsp;&lt;A href="https://www.nxp.com/design/software/development-software/eiq-ml-development-environment/eiq-inference-with-tensorflow-lite-micro:EIQ-TFLITE-MICRO" target="_self" rel="nofollow noreferrer"&gt;eIQ&lt;/A&gt;/&lt;/SPAN&gt;&lt;A href="https://www.tensorflow.org/lite/microcontrollers" target="_blank" rel="nofollow noopener noreferrer"&gt;TensorFlow Lite for Microcontrollers&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;(TFLM) is compatible as well.&lt;/P&gt;&lt;P&gt;What I want to do is design a bare-metal (no OS, so no Python etc.) Embedded System with a Cortex-M microcontroller and use the TPU to accelerate an image classifier using TFLM. I'd like to use the TPU either through the standalone chip (&lt;A href="https://coral.ai/products/accelerator-module" target="_self"&gt;Accelerator Module&lt;/A&gt;) or the&amp;nbsp;&lt;A href="https://coral.ai/products/accelerator/" target="_self"&gt;USB Accelerator&lt;/A&gt;. They both use USB2.0.&lt;/P&gt;&lt;P&gt;There's a similar system from Google, the&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://coral.ai/products/dev-board-micro/" target="_blank" rel="nofollow noopener noreferrer"&gt;Dev Board Micro&lt;/A&gt;, which mounts a Cortex-M (i.MX RT1170) and according to the product description "Supports TensorFlow Lite and TensorFlow Lite for Microcontrollers". But unfortunately it's still "coming soon" and I don't find any other useful info or similar projects online. Being an official product I assume TFLM (and as a result, eIQ) should support the Edge TPU but I don't understand whether it's already supported&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;EM&gt;now&lt;/EM&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;or maybe it will be in the future only when the Dev Board Micro is released.&lt;/P&gt;&lt;P&gt;I tried to have a look at the&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://github.com/tensorflow/tflite-micro/blob/main/tensorflow/lite/c/common.h" target="_blank" rel="nofollow noopener noreferrer"&gt;GitHub&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;repo of TFLM and at the line 56 I found this but I don't know to how to interpret it:&lt;/P&gt;&lt;BLOCKQUOTE&gt;&lt;P&gt;kTfLiteEdgeTpuContext = 2, // Placeholder for Edge TPU support.&lt;/P&gt;&lt;/BLOCKQUOTE&gt;&lt;P&gt;Thank you!&lt;/P&gt;</description>
    <pubDate>Thu, 14 Jul 2022 14:55:46 GMT</pubDate>
    <dc:creator>anonymous3957</dc:creator>
    <dc:date>2022-07-14T14:55:46Z</dc:date>
    <item>
      <title>Does eIQ support Google Edge TPU (through Accelerator Module or USB Accelerator)?</title>
      <link>https://community.nxp.com/t5/i-MX-RT-Crossover-MCUs/Does-eIQ-support-Google-Edge-TPU-through-Accelerator-Module-or/m-p/1489590#M20613</link>
      <description>&lt;P&gt;I already tried to ask &lt;A href="https://stackoverflow.com/questions/71590845/does-tensorflow-lite-for-microcontrollers-support-google-edge-tpu/71685429" target="_self"&gt;here&lt;/A&gt;, but didn't get a satisfying answer. I'll try again in here since I want to use the i.MX RT1170.&lt;/P&gt;&lt;P&gt;I already know that TensorFlow Lite (TFL)&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://coral.ai/docs/edgetpu/models-intro/" target="_blank" rel="nofollow noopener noreferrer"&gt;supports&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;the Google Edge TPU, for instance through the&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://coral.ai/products/dev-board/" target="_blank" rel="nofollow noopener noreferrer"&gt;Coral Dev Board&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;(Linux required).&lt;/P&gt;&lt;P&gt;However I'd like to know whether&lt;SPAN&gt;&amp;nbsp;&lt;A href="https://www.nxp.com/design/software/development-software/eiq-ml-development-environment/eiq-inference-with-tensorflow-lite-micro:EIQ-TFLITE-MICRO" target="_self" rel="nofollow noreferrer"&gt;eIQ&lt;/A&gt;/&lt;/SPAN&gt;&lt;A href="https://www.tensorflow.org/lite/microcontrollers" target="_blank" rel="nofollow noopener noreferrer"&gt;TensorFlow Lite for Microcontrollers&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;(TFLM) is compatible as well.&lt;/P&gt;&lt;P&gt;What I want to do is design a bare-metal (no OS, so no Python etc.) Embedded System with a Cortex-M microcontroller and use the TPU to accelerate an image classifier using TFLM. I'd like to use the TPU either through the standalone chip (&lt;A href="https://coral.ai/products/accelerator-module" target="_self"&gt;Accelerator Module&lt;/A&gt;) or the&amp;nbsp;&lt;A href="https://coral.ai/products/accelerator/" target="_self"&gt;USB Accelerator&lt;/A&gt;. They both use USB2.0.&lt;/P&gt;&lt;P&gt;There's a similar system from Google, the&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://coral.ai/products/dev-board-micro/" target="_blank" rel="nofollow noopener noreferrer"&gt;Dev Board Micro&lt;/A&gt;, which mounts a Cortex-M (i.MX RT1170) and according to the product description "Supports TensorFlow Lite and TensorFlow Lite for Microcontrollers". But unfortunately it's still "coming soon" and I don't find any other useful info or similar projects online. Being an official product I assume TFLM (and as a result, eIQ) should support the Edge TPU but I don't understand whether it's already supported&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;EM&gt;now&lt;/EM&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;or maybe it will be in the future only when the Dev Board Micro is released.&lt;/P&gt;&lt;P&gt;I tried to have a look at the&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;&lt;A href="https://github.com/tensorflow/tflite-micro/blob/main/tensorflow/lite/c/common.h" target="_blank" rel="nofollow noopener noreferrer"&gt;GitHub&lt;/A&gt;&lt;SPAN&gt;&amp;nbsp;&lt;/SPAN&gt;repo of TFLM and at the line 56 I found this but I don't know to how to interpret it:&lt;/P&gt;&lt;BLOCKQUOTE&gt;&lt;P&gt;kTfLiteEdgeTpuContext = 2, // Placeholder for Edge TPU support.&lt;/P&gt;&lt;/BLOCKQUOTE&gt;&lt;P&gt;Thank you!&lt;/P&gt;</description>
      <pubDate>Thu, 14 Jul 2022 14:55:46 GMT</pubDate>
      <guid>https://community.nxp.com/t5/i-MX-RT-Crossover-MCUs/Does-eIQ-support-Google-Edge-TPU-through-Accelerator-Module-or/m-p/1489590#M20613</guid>
      <dc:creator>anonymous3957</dc:creator>
      <dc:date>2022-07-14T14:55:46Z</dc:date>
    </item>
    <item>
      <title>Re: Does eIQ support Google Edge TPU (through Accelerator Module or USB Accelerator)?</title>
      <link>https://community.nxp.com/t5/i-MX-RT-Crossover-MCUs/Does-eIQ-support-Google-Edge-TPU-through-Accelerator-Module-or/m-p/1491901#M20696</link>
      <description>&lt;P&gt;Hi&amp;nbsp;&lt;a href="https://community.nxp.com/t5/user/viewprofilepage/user-id/203702"&gt;@anonymous3957&lt;/a&gt;,&lt;/P&gt;&lt;P&gt;This is the only official information we have available about TFLM as of now:&amp;nbsp;&lt;A href="https://www.nxp.com/design/software/development-software/eiq-ml-development-environment/eiq-inference-with-tensorflow-lite-micro:EIQ-TFLITE-MICRO" target="_blank"&gt;eIQ® Inference with TF Lite Micro | NXP Semiconductors.&amp;nbsp;&lt;/A&gt;&lt;/P&gt;&lt;P&gt;The i.MX RT 1170 is listed which means it is compatible with the software. Any other question related to Google TPU should be asked to them, unfortunately.&lt;/P&gt;&lt;P&gt;Hope this helps and best regards,&lt;/P&gt;&lt;P&gt;Julian&lt;/P&gt;</description>
      <pubDate>Tue, 19 Jul 2022 15:38:40 GMT</pubDate>
      <guid>https://community.nxp.com/t5/i-MX-RT-Crossover-MCUs/Does-eIQ-support-Google-Edge-TPU-through-Accelerator-Module-or/m-p/1491901#M20696</guid>
      <dc:creator>Julián_AragónM</dc:creator>
      <dc:date>2022-07-19T15:38:40Z</dc:date>
    </item>
  </channel>
</rss>

