Optimizing EmberGen Simulations for Large-Scale Scenes

Software Support & Suggestions
1

Hello

I’ve been experimenting with EmberGen for creating large-scale fire & smoke simulations; but I’m running into performance limitations when working with high-resolution settings. Even with a powerful GPU; the simulation slows down significantly, making it challenging to iterate quickly. Are there any recommended best practices for optimizing EmberGen projects without sacrificing too much detail? :frowning_face:

Additionally; I’m curious if there are any specific settings or workflow adjustments that can help improve the efficiency of rendering and exporting volumetric data for use in external applications like Unreal Engine or Blender. Are there compression techniques or voxel resolution tweaks that can help maintain quality while improving performance? :thinking: Checked EmberGen Documentation - Documentationazure solution architect resources, and found it quite informative.

I’d love to hear from others who have tackled similar challenges & any insights from the JangaFX team on upcoming performance improvements. Any official resources or documentation links would also be appreciated! :slightly_smiling_face:

Thank you !! :slightly_smiling_face:

2

Hi there, the best practice I can recommend is trying to find the maximum number of voxels you can export while still maintaining tolerable export times. I have found that with my 4080 my limit is around 1 billion active voxels (including upscaling), knowing that I will adjust my simulation bounds so that when using 3x upscaling I get around 1 billion active voxels. Part of this depends on the limitations of your graphics card so it requires some experimenting on your end. Let me know if this helps!