I passed multiple gpu devices in configuration, nvidia-smi shows multiple gpus in host and in docker. When I run your comfyui environment and I look at nvidia-smi I only see memory and gpu load active on 1 gpu. Is the software able to share gpu memory and gpu compute using multiple gpu?
I passed multiple gpu devices in configuration, nvidia-smi shows multiple gpus in host and in docker. When I run your comfyui environment and I look at nvidia-smi I only see memory and gpu load active on 1 gpu. Is the software able to share gpu memory and gpu compute using multiple gpu?
Thank you