search

google / CFU-Playground

Want a faster ML processor? Do it yourself! -- A framework for playing with custom opcodes to accelerate TensorFlow Lite for Microcontrollers (TFLM). . . . . . Online tutorial: https://google.github.io/CFU-Playground/ For reference docs, see the link below.
http://cfu-playground.rtfd.io/
Apache License 2.0
455 stars 116 forks source link

Enable any LiteX-supported board to be used. #85

Open tcal-x opened 3 years ago

tcal-x commented 3 years ago

CFU-Playground currently supports Arty 35T and Nexys Video boards. The difference between them is very small, but the current setup requires duplication of at least a couple of files in the soc/ directory (arty.mk, arty.py, nexys_video.mk, nexys_video.py). Instead, the board should be a parameter so that just one .mk and one .py file are needed.

Of course, not all LiteX-supported boards have the fabric resources to handle a "full" VexRiscv variant, or the memory to handle both the loaded executable and the necessary work space (e.g. for the tensor arena), so we may want to perform some up-front checks.

PiotrZierhoffer commented 3 years ago

@tcal-x should we close this issue?