So, I have a few RTX gpus in a server.
I use it for folding at home.
Moo!wrapper uses very little resources of the GPU, so I wanted to share the GPU with FAH, to keep all GPUs 100% under load, while still supporting both projects.
It works fine, except for Moo!Wrapper.
Moo Wrapper only can process 1 GPU at a time, as with 2 GPUs (I'm running CUDA on UBUNTU), for some odd reason, the tasks stall.
Milkyway, Einstein, all seem to do fine with running multi gpus on this setup.

So I went into the project, and created an app_config.xml file with following contents:

<app_config>
   [<app>
      <name>Moo! Wrapper</name>
      <max_concurrent>1</max_concurrent>
      </gpu_versions>
    </app>]
   <project_max_concurrent>1</project_max_concurrent>
</app_config>

However, Moo! Wrapper doesn't honor that request of running only 1 GPU (out of 4).

Help me out here!
I do want to support moo! Wrapper.