2019-09-06 12:33 PM
Hello,
I have a question concerning the memory footprint of a neural network.
Given a pre-trained neural network topology, how can I estimate the needed memory footprint.
Are there any application notes that describes this in more detail.
All feedback is welcome.
Best regards,
Joeri
2019-09-09 12:06 AM
Hi,
When you use X-Cube-AI and analyze the network it will report the amount of Flash and Ram needed to run the network
Also the tool automatically computes the stack size and heap size needed.
Regards
Daniel
2019-09-09 09:47 AM
Daniel,
Thanks.
In meantime I was checking the X-Cube-AI tool.
Best regards,
Joeri
2019-09-11 08:21 AM
I notice that PyTorch is not in the list of supported deep learning frameworks.
Correct?
Is this on the product roadmap.
2019-09-14 01:56 PM
I found my Answer.
PyTorch is on the roadmap for 2020.
Perfect!