Hi, there is any possibility of using Jetson Orin Devkit DLA cores during Holoscan’s TensorRT inference?
I checked Jtop and DLA cores are not used by default. If i use them i can have an improve of performance. Is there any documentation on how to do this?
Thanks