Jetson's Nano GPU as a GPU device
4 Ansichten (letzte 30 Tage)
Ältere Kommentare anzeigen
I have managed to successfully connect with my NVidia Jetson Nano 2GB board. I can see from the connection details, that a GPU of the Jetson is NVIDIA Tegra X1.
I would like to use this GPU for deep learning network training (as in parallel computation). I thought that when I successfully connect with Jetson, Matlab will detect available GPU and set it as my gpuDevice.
However, when I type in gpuDevice, it returns error:
The CUDA driver could not be loaded. The library name used was 'nvcuda.dll'.
The specified module could not be found.
I have downloaded and installed CUDA Toolkit. I tried searching for NVIDIA Tegra X1 drivers to install on my windows system, but can't find them. Is it even possible to run Matlab computation using GPU embedded on Jetson board?
2 Kommentare
Nathan Malimban
am 16 Feb. 2022
Hi Jagoda,
When you installed the toolkit from NVIDIA's website, did you also install the driver? I believe the installation gives you the option to install one, or the other, or both. It's possible you only installed the toolkit.
You can also take a look at the answer to this MATLAB Answers post, as it looks similar:
https://www.mathworks.com/matlabcentral/answers/501461-installed-cuda-but-unable-to-run-gpuarrays
Antworten (1)
Ramakrishna Mandalapu
am 20 Jan. 2023
Hi Jagoda,
Jetson Nano's GPU is outside of MATLAB environment, so you can not use this as the one on your host (where MATLAB runs). Though you can excercise various workflows on this board using GPU Coder and MATLAB Coder Support Package for NVIDIA Jetson and NVIDIA DRIVE Platforms.
0 Kommentare
Siehe auch
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!