# Lambda Stack version archive

**URL:** <https://deeptalk.lambda.ai/t/lambda-stack-version-archive/3727>\
**Category:** Technical Help\
**Created:** [May 25, 2023, 3:27pm UTC](https://deeptalk.lambda.ai/t/lambda-stack-version-archive/3727 "2023-05-25T15:27:41Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![daryl149](https://avatars.discourse-cdn.com/v4/letter/d/b487fb/32.png) [@daryl149](https://deeptalk.lambda.ai/u/daryl149)\
**Post date:** [May 25, 2023, 3:27pm UTC](https://deeptalk.lambda.ai/t/lambda-stack-version-archive/3727/1 "2023-05-25T15:27:41Z")

</div>

Hi, is there any way to install historical versions of the lambda stack? I am troubleshooting the latest install (which comes with CUDA 12.0) for something that works flawlessly on my machine running an earlier version of the lambda stack (that came with CUDA 11.6)

---

<div class="post-metadata">

**Author:** ![corey](https://avatars.discourse-cdn.com/v4/letter/c/838e76/32.png) [@corey](https://deeptalk.lambda.ai/u/corey)\
**Post date:** [May 30, 2023, 1:35pm UTC](https://deeptalk.lambda.ai/t/lambda-stack-version-archive/3727/2 "2023-05-30T13:35:28Z")

</div>

There are not currently historical versions of Lambda Stack available. For help with any Lambda Stack issues, please open a ticket at the following url.

[https://support.lambdalabs.com/hc/en-us/requests/new](https://support.lambdalabs.com/hc/en-us/requests/new)

---

<div class="post-metadata">

**Author:** ![markd](https://avatars.discourse-cdn.com/v4/letter/m/839c29/32.png) [@markd](https://deeptalk.lambda.ai/u/markd)\
**Post date:** [June 14, 2023, 2:41am UTC](https://deeptalk.lambda.ai/t/lambda-stack-version-archive/3727/3 "2023-06-14T02:41:54Z")

</div>

The latest Lambda Stack comes with 525.116.04, which supports _up to_ CUDA 12.0.  
The current CUDA in use is 11.8.

The Driver version is commonly newer than the CUDA version, as most issues are in CUDA versus the driver and the driver supports a few major versions back.

You can check the CUDA version typically with:  
$ nvcc --version  
Or from PyTorch:  
$ python -c ‘import torch ; print("\nIs available: ", torch.cuda.is\_available()) ; print("Pytorch CUDA Compiled version: ", torch.\_C.\_cuda\_getCompiledVersion()) ; print("Pytorch version: ", torch. **version** ) ; print("pytorch file: ", torch. **file** ) ; num\_of\_gpus = torch.cuda.device\_count(); print("Number of GPUs: ",num\_of\_gpus)’

To version you CUDA you can use a virtual environment or docker container.  
I have a number of examples at:  
[lambdalabs/documentation/software/examples/virtual-environments at main · markwdalton/lambdalabs · GitHub](https://github.com/markwdalton/lambdalabs/tree/main/documentation/software/examples/virtual-environments)

We should have better write up soon at ‘[http://docs.lambdalabs.com/](http://docs.lambdalabs.com/)’