Is it possible to utilize multiple GPUs for rendering out of sequencer to speed up render times? I see conflicting information online, and in the task manager only 1 GPU is being utilized. Using 5.2.
Our machine has 4x 4090 cards and work without issue in other programs such as Blender.
The path tracer can optionally use multiple GPUs if they are connected with SLI (which doesn’t apply to the 4090), but the regular real-time renderer doesn’t have any way of doing so. As far as rendering sequences out of MRQ, in theory it is possible to write an MRQ pipeline executor to split the sequence up so that each shot is rendered separately on four Unreal instances running on the machine, each assigned a different GPU, but this would take a bit of work to write and also you would need enough CPU cores and memory to handle four running UE instances.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.