# How to check torch gpu compatibility without initializing CUDA?

**URL:** <https://discuss.pytorch.org/t/how-to-check-torch-gpu-compatibility-without-initializing-cuda/128528>\
**Category:** deployment\
**Created:** [August 4, 2021, 11:54am UTC](https://discuss.pytorch.org/t/how-to-check-torch-gpu-compatibility-without-initializing-cuda/128528 "2021-08-04T11:54:10Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![Manuel\_Alejandro\_Dia](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/manuel_alejandro_dia/32/35111_2.png) [@Manuel\_Alejandro\_Dia](https://discuss.pytorch.org/u/Manuel_Alejandro_Dia)\
**Post date:** [August 4, 2021, 12:21pm UTC](https://discuss.pytorch.org/t/how-to-check-torch-gpu-compatibility-without-initializing-cuda/128528/4 "2021-08-04T12:21:07Z")

</div>

It seems like you are correct assuming:

> Since the answer here might not be very torch-related and instead implicate usage of e.g. nvidia specific tools

According to [this comment](https://discuss.pytorch.org/t/solved-pytorch-no-longer-supports-this-gpu-because-it-is-too-old/15444/10) there is a relationship between the CUDA version and the deprecated compute capabilities you won’t be able to use. You can check the [CUDA 10.2 Toolkit Docs](https://docs.nvidia.com/cuda/archive/10.2/cuda-toolkit-release-notes/index.html#deprecated-features) where it says which [compute capabilities](https://developer.nvidia.com/cuda-gpus) are deprecated.

---

_[View the full topic](https://discuss.pytorch.org/t/how-to-check-torch-gpu-compatibility-without-initializing-cuda/128528)._