Is it possible to do an Inference on an Artix7 esp. a Digilent Nexys Video Board?

Hello,

i do have a Digilent Nexys Video Board with a Artix7 FPGA. Im trying do deploy a NN on it with the help of MATLAB Deep Learning HDL Toolbox.

I followed this tutorial: https://de.mathworks.com/help/deep-learning-hdl/ug/define-custom-board-and-reference-design-for-dl-ip-core-workflow.html

I registered my Board and Reference Design.

But im getting a Error Message as soon as i try to "dlhdl.buildProcessor(hPC, 'ProjectFolder','genericipcore_prj','ProcessorName','GenericProcessor','HDLCoderConfig',{'TargetLanguage','VHDL'});".

The Message is as follows: Error using dlhdl.buildProcessor

Device family not supported for Resource estimation. Refer to the documentation for steps to register a new device for

Area Estimation.

Error in CustomIPCoreGen (line 13)

dlhdl.buildProcessor(hPC, 'ProjectFolder','genericipcore_prj','ProcessorName','GenericProcessor','HDLCoderConfig',{'TargetLanguage','VHDL'});

Kind regards

Silas

Answer 1

[Tom Richter]

Hi Silas,

I suggest contacting Technical Support to help you with this issue/question. If you go to https://www.mathworks.com/support/contact_us.html, you can submit a support request and add example or reproduction files. The page also displays phone contact information based on your location.

Notice that the Artix7 you use only has 1/3 of the hardware resources that the KCU105 or ZCU102 have. A neural network with convolution layers might not perform fast enough. You would definitely need to quantize the network before implementation.

Best regards,

Tom