Block-Structured Cartesian Cut-Cell AMR for Production RANS and LES
Traditional CFD meshing was designed for CPU-era solvers - manual, geometry-dependent, and slow. Our white paper introduces a fundamentally different approach: a block-structured Cartesian cut-cell method with Adaptive Mesh Refinement that's built from the ground up for GPU execution. We eliminate manual cleanup and enable fully automated, high-quality grid generation for both RANS and LES - with anisotropic refinement that captures boundary layers and critical flow features without wasting cells in regions that don't need them.
The regular Cartesian structure maps naturally onto GPU memory hierarchies and parallel compute patterns, unlocking simulation speeds that unstructured meshes simply can't match. This is the meshing architecture that makes real-time Physics AI feasible.
Access the Whitepaper
Enter your email below to download the full technical whitepaper.
