Hi Gabe,
Thanks for the file.
The bottleneck is in the Multi-Group. See the analytics below with 15 million particles, which are an added workload. I'm unclear if the Multi Group pulls the data from the GPU back into the CPU here. Perhaps test your scene to see if Caching will help. My laptop here will not show your performance, I assume.
Screenshot 2024-11-02 at 10.27.58 PM.jpg
I have copied the Color Mapper and the Advection to each Group. Please take a look. I do not have the capacity here to render many tests. Sorry.
Enjoy your weekend