# Upgrade to latest CUDA on Jetson without upgrading JetPack **URL:** \ **Category:** Jetson AGX Orin\ **Created:** [October 13, 2022, 2:30am UTC](https://forums.developer.nvidia.com/t/upgrade-to-latest-cuda-on-jetson-without-upgrading-jetpack/230607 "2022-10-13T02:30:58Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![kayccc](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/kayccc/32/14043_2.png) [@kayccc](https://forums.developer.nvidia.com/u/kayccc)\ **Post date:** [October 13, 2022, 2:30am UTC](https://forums.developer.nvidia.com/t/upgrade-to-latest-cuda-on-jetson-without-upgrading-jetpack/230607/1 "2022-10-13T02:30:59Z")
_Starting from CUDA 11.8, Jetson users on JetPack 5.0 and above will have the ability to upgrade to the latest and greatest CUDA versions without the need to update the JetPack version or Jetson Linux BSP._ _To know more on how to upgrade CUDA on your Jetson device today, read our blog [Simplifying CUDA Upgrades for NVIDIA Jetson Users | NVIDIA Technical Blog](https://developer.nvidia.com/blog/simplifying-cuda-upgrades-for-nvidia-jetson-users/)_ - _On the_ [_CUDA 11.8 Downloads_](https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdeveloper.nvidia.com%2Fcuda-downloads&data=05%7C01%7Csuhash%40nvidia.com%7Cd8e5ebfd10c64901731808da94d7d4ee%7C43083d15727340c1b7db39efd9ccc17a%7C0%7C0%7C637985951956451491%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=JJc6WOS%2B6KgO9%2BFeHqu3T%2F6BnhVmVWN%2F2A8Uxqwtx8E%3D&reserved=0) _page, download the CUDA installer for “aarch64-Jetson” and follow the installation instructions to upgrade your Jetson device to CUDA 11.8._ - _For more details on CUDA upgradable package on Jetson, refer to the_ [_CUDA for Tegra App Note_](https://nam11.safelinks.protection.outlook.com/?url=http%3A%2F%2Fcuda-internal.nvidia.com%2Frelease-candidates%2Fkitpicks%2Fcuda-r11-8%2F11.8.0%2F029%2Fdocs%2Fportal%2Fcuda-for-tegra-appnote%2Findex.html%23upgradable-package-for-jetson&data=05%7C01%7Csuhash%40nvidia.com%7Cd8e5ebfd10c64901731808da94d7d4ee%7C43083d15727340c1b7db39efd9ccc17a%7C0%7C0%7C637985951956451491%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=KhtK%2Bax%2BauRf7wrK%2Bv6ysQMibTW173axyS5PZ5AUcKk%3D&reserved=0)_._ - _For information on all the new features that CUDA 11.8 brings in, refer to the_ [_release notes_](https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html#title-new-features)_._ - _Do register for our deep-dive webinar_ [_here_](https://gateway.on24.com/wcc/experience/elitenvidiabrill/1407606/3944300/jetpack-502) _where the CUDA and Jetson teams will walk you through details on this new feature and get an opportunity to ask questions live! “_ ---
**Author:** ![kayccc](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/kayccc/32/14043_2.png) [@kayccc](https://forums.developer.nvidia.com/u/kayccc)\ **Post date:** [October 13, 2022, 2:31am UTC](https://forums.developer.nvidia.com/t/upgrade-to-latest-cuda-on-jetson-without-upgrading-jetpack/230607/2 "2022-10-13T02:31:07Z")
Pinned on October 13, 2022, 2:31am UTC ---
**Author:** ![dusty\_nv](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/dusty_nv/32/14043_2.png) [@dusty\_nv](https://forums.developer.nvidia.com/u/dusty_nv)\ **Post date:** [June 6, 2023, 2:47am UTC](https://forums.developer.nvidia.com/t/upgrade-to-latest-cuda-on-jetson-without-upgrading-jetpack/230607/4 "2023-06-06T02:47:39Z")
Unpinned on June 6, 2023, 2:47am UTC ---
**Author:** ![johnnynunez](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/johnnynunez/32/385239_2.png) [@johnnynunez](https://forums.developer.nvidia.com/u/johnnynunez)\ **Post date:** [September 29, 2023, 3:19pm UTC](https://forums.developer.nvidia.com/t/upgrade-to-latest-cuda-on-jetson-without-upgrading-jetpack/230607/5 "2023-09-29T15:19:37Z")
Now with cuda 12.2. Driver is too old… > > > import torch > > > torch.cuda.is\_avaible() > > > Traceback (most recent call last): > > > File “”, line 1, in > > > AttributeError: module ‘torch.cuda’ has no attribute ‘is\_avaible’. Did you mean: ‘is\_available’? > > > torch.cuda.is\_available() > > > /home/jetson/miniconda3/envs/prueba/lib/python3.11/site-packages/torch/cuda/ **init**.py:138: UserWarning: CUDA initialization: The NVIDIA driver on your system is too old (found version 11040). Please update your GPU driver by downloading and installing a new version from the URL: [Official Drivers | NVIDIA](http://www.nvidia.com/Download/index.aspx) Alternatively, go to: [https://pytorch.org](https://pytorch.org) to install a PyTorch version that has been compiled with your version of the CUDA driver. (Triggered internally at /home/jetson/Projects/pytorch/c10/cuda/CUDAFunctions.cpp:108.) > > > return torch.\_C.\_cuda\_getDeviceCount() \> 0 > > > False