Can this SDK be used to run Deep learning models on Adreno 330 GPU with Snapdragon 801, Quad-core 2.26 GHz quad-core Krait CPU?
The NPE SDK supports Snapdragon® 820, 835, 625, 626, 650, 652, 653, 660, 630 and 450 as well as the Snapdragon® 820Am automotive platform. For Qualcomm® Adreno™ GPU support, libOpenCL.so must be present on device.
Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries (“Qualcomm”). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.
Hi. Thanks for the question.
Unfortunately, the 801 is not an officially supported platform. However, the SNPE CPU runtime should, at least in theory, work on any Snapdragon CPU. How fast it runs is another question. For unsupported Adreno GPU versions, whether SNPE works at all, and if it does works, how fast it runs depends on a several factors (e.g. the version of OpenCL present on the device, what the particular network you're trying to run looks like as some layers might work and others might not, etc.). You are, of course, welcome to try it out on the 801, but you're on your own as far as any issues you encounter on an unsupported platform.