How to to install cuda 10.0 on jetson nano separately ?


I have a sdcard.img file with Jetpack SDK 4.2. I was able to flash into memory card and run it on jetson nano.

The issue here was - no Cuda installed in that image.
I am following this article to install it.
I need .deb files to install it on the jetson nano. I was unable to find .deb files online. So I am using Nvidia SDK manager to download the necessary .deb files into a local ubuntu system.
Then, I copied those downloaded files into the sd card which was flashed with Jetpack SDK 4.2 and then installing them on starting jetson nano.

I am not sure whether this procedure was correct or wrong? Did anyone face this issue


1 Like


CUDA and all the packages are installed by the JetPack installer.
After setting up the SD card, you should also run the SDK manager to flash/install the latest OS/packages into the Nano.


1 Like

SDK manager is not available on my Nano. I am using the image provided in Getting Started (Jetson Nano Developer Kit).

I downloaded the SDK deb file to install on my Nano. It shows 0 bytes on the installer and does not work.

Also, there is no /user/local/cuda folder.

Please advise. Thank you!

1 Like

You can use the OTA updates rather than SDK Manager to install anything as of JetPack 4.3. Before that release, you had to use a Linux desktop and SDK Manager, unfortunately.

sudo apt install cuda-core-10-0

Will install cuda itself, but it should already be installed on the default rootfs. If not (apt will tell you if it’s already installed or not), you must install nvidia’s apt key (available in the BSP tarball under the Linux_for_Tegra/nv_tegra/ folder) and then add teh apt sources to a sources.list. For nano the lines should look like this:

(from /etc/apt/sources.list.d/nvidia-l4t-apt-source.list)

deb r32 main
deb r32 main

Some other packages:

$ sudo apt install cuda-
cuda-command-line-tools-10-0        cuda-curand-dev-10-0                cuda-license-10-0                   cuda-nvprof-10-0
cuda-compiler-10-0                  cuda-cusolver-10-0                  cuda-memcheck-10-0                  cuda-nvprune-10-0
cuda-core-10-0                      cuda-cusolver-dev-10-0              cuda-minimal-build-10-0             cuda-nvrtc-10-0
cuda-cublas-10-0                    cuda-cusparse-10-0                  cuda-misc-headers-10-0              cuda-nvrtc-dev-10-0
cuda-cublas-dev-10-0                cuda-cusparse-dev-10-0              cuda-npp-10-0                       cuda-nvtx-10-0
cuda-cudart-10-0                    cuda-documentation-10-0             cuda-npp-dev-10-0                   cuda-repo-l4t-10-0-local-10.0.326
cuda-cudart-dev-10-0                cuda-driver-dev-10-0                cuda-nsight-compute-addon-l4t-10-0  cuda-samples-10-0
cuda-cufft-10-0                     cuda-gdb-10-0                       cuda-nvcc-10-0                      cuda-toolkit-10-0
cuda-cufft-dev-10-0                 cuda-gdb-src-10-0                   cuda-nvdisasm-10-0                  cuda-tools-10-0
cuda-cuobjdump-10-0                 cuda-gpu-library-advisor-10-0       cuda-nvgraph-10-0                   
cuda-cupti-10-0                     cuda-libraries-10-0                 cuda-nvgraph-dev-10-0               
cuda-curand-10-0                    cuda-libraries-dev-10-0             cuda-nvml-dev-10-0

You can use tab completion or “apt search” to look for specific package names or keywords, respectively.

Most of this is intalled on the rootfs by default, but if it’s not, you can apt-get what you want now.

1 Like

you may mean /usr/local/cuda ?

1 Like


That should exist on the default rootfs provided. If it doesn’t, something is probably very wrong and you may wish to just reflash JetPack 4.3 on a new SD card. SDK Manager is meant to be installed on a Linux Desktop and cannot be installed on the Nano itself (However there is no longer any real need for it at all, at least with the Dev kit).

I just used another microSD card and downloaded the Jetpack 4.3 image from to see if that would make a difference - it didn’t.

1 Like

What does this command show when you run it on your nano?

apt list --installed | grep cuda

how about?

ls /usr/local/cuda

It shows CUDA 10.0.326-1 installed, automatic (unknown, stable, now). I must mention that I’m not new, but VERY rusty with programming. Thanks for helping me.

1 Like

No worries. Glad to help. That’s what the forum is for. Feel free to ask for any examples in a new thread or search the forum with Google.


What about cudnn??

I flashed the image in jetson nano through baleno etcher. It shows cuda 10.0, but shows cudnn 7.5.
I want the cudnn rather 7.6 ver. How can i upgrade it?

Thanks in advance
Abhi J K

Same. they can’t be upgraded individually, only through JetPack with the version included.

I currently have cuda10.2, so what exactly do you suggest for downgrading it

My understanding of what @kayccc is saying is the versions are tied to the JetPack. I’m not aware of any way currently to downgrade the JetPack other than backing up and reflashing, but someone else might suggest a possible path.

I also flashed the same image onto my Nano via SD and it also installed CUDA 10.2.
I installed CUDA 10.0.0 via apt install as @mdegans pointed out. This was enough for me to be able to successfully use the version of PyTorch installed by the script that comes with the SD image from the “Getting Started” page. So, while you can’t “downgrade”, you can have 10.0.0 installed parallel to 10.2 .

It would be useful and save time for new users if SDKs that use CUDA were built to use the latest version of CUDA supplied by the flashable image.

It may work, but be aware that the configuration is unsupported for downgrading. I did notice it’s possible to have parallel versions of some libraries since some offline apt repos were still in the lists on my Xavier, but I expect some side effects.

If you notice any unusual behavior, I would backup and start from scratch. I really don’t think Nvidia tested it like this. I was watching the other thread and @dusty_nv is right that PyTorch should be building. It may be there is something else going on. A reflash never hurts to rule things out.

I’ve reflashed twice just to make sure that I wasn’t missing something or didn’t pull a package down from a different repo. I’ve retraced my steps through the “Getting Started” and "Hello AI World " tutorials and no, Pytorch fails to install from the installer script run here:

It looks like torch-1.1.0-cp36-cp36m-linux_aarch64.whl that is downloaded is not built against CUDA 10.2 and is failing looking for 10.0.0

@dusty_nv you may want to have a look at this.

It looks like the script you refer to might need to be updated. Try downloading 1.5 from here for the latest JetPack:

To be honest I don’t know a lot about PyTorch, but my understanding is 1.5 is for the latest JetPack with CUDA 10.2

Do you need a specific version? If so it might be worth it to reflash with 4.3, and only upgrade when the thing you need is ready for 4.4. Upgrading should work without a problem, but downgrading… Well it’s untested. Apt itself works pretty well, but the packages themselves probably weren’t designed for this.

Hi all, the script has been updated to install the PyTorch wheels for CUDA 10.2, sorry about that. Please refer to this post for more information:

It is not recommended or necessary to downgrade to CUDA 10.0, please stay with CUDA 10.2.