Is it possible to use different target hardware for implementing deep learning HDL toolbox?

Hi,

Currently, i am exploring about the implementation of deep learning HDL toolbox. I looked the web documentation and several youtube videos for it, but i am still confused for hardware target.

I have FPGA with the model of Arty-Z7 and XUP Pynq-ZU, but there are no available example for it and the support hardware type does not include it in the model. Is it still possible to implement deep learning HDL toolbox for those hardware?

If we generate the HDL from deep learning HDL toolbox like in this example https://www.youtube.com/watch?v=cwbJSBhBUks , and if we use the different target hardware but in the same family chip, does it work for those hardware?

If anyone has related experience about implementation deep learning algorithm in FPGA using MATLAB, i would love to read your experience in the comment below. Thank you very much.

Answer 1

[Kiran Kintali]

DL HDL ships bitstreams for few reference boards.

However, DL HDL IP can be customized to any custom FPGA / ASIC configuration.

https://www.mathworks.com/help/deep-learning-hdl/dlarchitecture.html

Answer 2

[Sai]

Hi Muhammad,

I understand that you want to use the Deep Learning HDL Toolbox on Arty-Z7 and XUP Pynq-ZU board.

As of R2022b, MathWorks Deep Learning HDL Toolbox does not support Arty-Z7 and XUP Pynq-ZU. Only the following boards are supported:

Xillinx:

• KCU105
• ZC706
• ZCU102

Intel:

• Arria10Soc

Please refer to the documentation for more information on Deep Learning HDL Toolbox supported hardware

[Muhammad Ihsan Al Hafiz]

Hi Sai,

Thank you for your answer.

I found this article https://se.mathworks.com/help/deep-learning-hdl/ug/define-custom-board-and-reference-design-for-dl-ip-core-workflow.html. Based on that article, if we generate the ip core, can we just use the supported board? or is it possible to create the custom board based on my hardware?

Thank you.