Jetson Orin: Running DLA and GPU cores at the same time

[shaumik.ganguly]

Is this old topic from Jetson Xavier still valid for Jetson Orin?

DLA and GPU cores at the same time

Hi, Here is a sample to run GPU and DLAs at the same time. 1. Please prepare TensorRT engine of GPU and DLA with trtexec first. For example, $ /usr/src/tensorrt/bin/trtexec --onnx=/usr/src/tensorrt/data/mnist/mnist.onnx --saveEngine=gpu.engine $ /usr/src/tensorrt/bin/trtexec --onnx=/usr/src/tensorrt/data/mnist/mnist.onnx --useDLACore=0 --allowGPUFallback --saveEngine=dla.engine 2. Compile TensorRT_sample.zip (3.5 KB) $ unzip TensorRT_sample.zip $ cd TensorRT_sample $ make 3. Test Please…

[AastaLLL]

Hi,

You can run trtexec in the different consoles directly.
Since there are some APIs changes in TensorRT, the sample (TensorRT_sample.zip) might not work on JetPack 5.

Thanks.

[shaumik.ganguly]

Hi AastaLLL,
Thanks!
I could however compile TensorRT_Sample with a modification to line48 in “tools.h” (it compiles successfully with some Warnings “TRT Deprecated” in NvInfer.h)

This is the change I made:
line 48: void log(Severity severity, const char* msg) override
to
line48: virtual void log(Severity severity, const char* msg) noexcept override

[AastaLLL]

Hi,

Thanks for the update.
It’s good to know the sample still works with the update.