Subject: High-performance LBM/SDF Engine (Hypercube Neo) seeking Dev Support"

Hi everyone, I’m working on Hypercube Neo, an MIT-licensed spatial computing fabric for the web.

I’ve reached a performance milestone of 262 MLUPS (1M cells updated in 4ms) on consumer-grade hardware. The engine uses a dual-topology approach:

  1. CPU: SharedArrayBuffer with bit-aligned strides to bypass V8 GC overhead.

  2. GPU: WebGPU compute shaders with $O(\log N)$ Jump Flooding for SDF.

I’m currently hitting Gemini limits while performing whole-repo reasoning for deep memory optimizations. If any Google/Chrome DevRel is around, I really love to discuss. Is there a research plan or free access for nice repo ? I just can’t do more than the individual pro plan using Antigravity

GitHub: https://github.com/Helron1977/Hypercube-neo/"