Hello, my name is Anton.
I’ve been working for a while on my SPH-based airfoil simulation based on CUDA. Generally, I’m using Tero Karras’ approach to find collisions between particles as fast as possible.
I’ve got some promising results. Particularly I can process up to 2’097’152 particles in real-time in 2D on 1080 TI at ~12 FPS. The physics part itself is not fully complete yet, but it’s not the most performance-critical part of the simulation to be focused on right now.
I’d like to share either the current or further results with the community. Is it possible to publish a post or an article on devblogs.nvidia.com in the future?
The Tero Karras’ approach:
https://devblogs.nvidia.com/wp-content/uploads/2012/11/karras2012hpg_paper.pdf (original paper)
Thinking Parallel, Part III: Tree Construction on the GPU | NVIDIA Technical Blog (brief explanation)
The simulation video: