Is GPU enabled or disabled in Jetson nano by default

Robotics & Edge Computing Jetson & Embedded Systems Jetson Nano
,
1

Does the GPU in the Jetson Nano need to be enabled manually, or does it activate automatically? If the GPU was not enabled, could that explain the slow response time while running an LLM like Llama? I’d appreciate it if you could provide an answer.

3

Hi,

GPU will be used if your code has CUDA support.

For example, if you run a model with PyTorch that has CUDA support.
The tasks will automatically run on the GPU.

To check if your tasks use GPU or not, please run tegrastats at the same to see if GPU resources (GR3D) are used.

$ sudo tegrastats

Thanks.

4

ok but I am using a Jetson Nano with the latest JetPack 4.6, which includes CUDA 10.2. It has come to my attention that PyTorch with GPU support for CUDA 10.2 is now outdated. Since the latest JetPack available is 4.6, it seems that I have no option left to utilize GPU with PyTorch on my Jetson Nano

5

Hi,

You can find some working packages in the below link:

Thanks.

7

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.