TurtLES benchmark released for turbulent 3D neural PDE surrogates

Researchers have introduced TurtLES, a substantial 14 TB dataset designed to serve as a benchmark for turbulent 3D neural PDE surrogates. This dataset comprises 400 trajectories of 3D turbulent cylinder wakes, computed using high-fidelity simulations, with each trajectory containing 400 time steps and millions of points per frame. TurtLES aims to address the limitations of existing benchmarks by incorporating physically realistic turbulence, dense irregular meshes, and long-horizon temporal dynamics, providing a more rigorous test for neural operators in computational fluid dynamics. AI

IMPACT Provides a new, large-scale benchmark for evaluating AI models in complex fluid dynamics simulations.

RANK_REASON The item describes a new benchmark dataset for evaluating neural PDE surrogates, which falls under research. [lever_c_demoted from research: ic=1 ai=1.0]

Read on Hugging Face Daily Papers →

• 2611.09567
• Angle of Attack
• computational fluid dynamics
• Hugging Face
• Implicit large eddy simulations of anisotropic weakly compressible turbulence with application to core-collapse supernovae
• Neural Operators
• Neural PDE Surrogates
• Reynolds number
• turbulent flow
• TurtLES
• Vortex stretching

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →