Can I use CUDA in containers on Windows 10 / GTS 450?

Accelerated Computing CUDA CUDA on Windows Subsystem for Linux
1

Is it impossible to use GTS 450 card by this thread way - Windows - Docker - WSL?

I have Windows 10 Pro 22H2 and installed CUDA toolkit in my WSL 2 together with nvidia-smi. I have latest available NVidia Windows driver version.
Running nvidia-smi in WSL gives me same error as when trying to run nvidia-smi in Docker containers:
“NVIDIA-SMI has failed because it couldn’t communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.”
nvidia-smi help tells that is supports architectures starting from Kepler while GTS 450 is older Fermi architecture. Same time lspci running in WSL shows no NVidia device…

So am I right that this GTS 450 will not work and it is hardware problem, not misconfiguration problem?
P.S. Generally GTS 450 is CUDA 2.1 …

2

WSL docs here state Pascal is the oldest supported architecture.

3

So looks like there is no way and I have to upgrade this ancient card first…

4

Yes, if you wish to follow the WSL route. Otherwise you can use the card on a suitable Linux distribution directly, Cuda Toolkit 8.0 is the last version to support CC 2.1.