Implementation of GPU and NPU usage / performance monitoring

[iska]

Hi @Andrey Ignatov ,

I'm using the NNAPI delegate of Tensorflow Lite to run our net. I am trying to determine which accelerator is used during runtime.
In the Burnout App there are three graphs that show the usage of CPU, GPU and NPU.
I know how to read the CPU usage, but I can not find a way to read the GPU or NPU usage programmatically.

I was wondering, how you managed to show the usages of all accelerators. Could you help me out with this?

Thank you in advance!
iska

 

[Andrey Ignatov]

I am trying to determine which accelerator is used during runtime.

You can sometimes find this information in the logcat when initializing the model with TFLite/NNAPI. Or you can tell us the device model you are using.

I can not find a way to read the GPU or NPU usage programmatically

There is no easy way to do this.