How much data can STAR-CCM+ collect from within a container?

Hi all,

I’ve got an entirely open source (bar some nvidia drivers) setup on my main linux system. However, I need to run starccm+ for some work stuff.

Currently I’m doing that in a virtual machine to isolate it from the main system. However, because the VM doesn’t have GPU passthrough for my nvidia card the graphics are quite laggy. As I understand it nvidia GPU passthrough to a VM is a complete nightmare and it would also leave my main system short a GPU. However, I have worked with nvidia GPU passthrough to a container in the past and that wasn’t too painful and worked well.

Therefore, I’m thinking of insteady trying to run starccm within a container on my main system by passing through the nvidia GPU and wayland socket. However, I’m don’t want starccm to be able to hover up a tonne of data on me, assuming I don’t give it file access to the host system are there any major privacy risks from running it in the container vs VM.

Thanks

You would be better off dedicating an entire hardware setup just for your professional workload while keeping your daily driver separate from it.

Ideally yes but that would be a very expensive solution. So I’m trying to work out if a container would work.

It is suboptimal, you would be trading security for convenience and usabiity.