{"id":"arxiv/2605.26388","url":"https://arxiv.org/abs/2605.26388","tags":["physics.comp-ph","math-ph"],"score":0,"title":"MARUT: An Exascale-Ready, GPU-Accelerated High-Order CFD Framework with AMR for High-Speed Flows and Finite-Rate Chemistry","domain":"physics.comp-ph","source":"arxiv","authors":["Trishit Mondal","Ameya D. Jagtap"],"reviews":0,"subject":"physics.comp-ph","summary":"We present MARUT, a scalable multi-GPU computational fluid dynamics (CFD) framework designed for high-fidelity simulations of compressible flows spanning subsonic to hypersonic regimes, including chemically reacting nonequilibrium flows with finite-rate chemistry and adaptive mesh refinement (AMR). The framework addresses a central challenge in contemporary scientific computing: the development of numerically accurate and computationally scalable algorithms capable of resolving strongly nonlinear, multiscale flow physics on emerging heterogeneous supercomputing architectures. Built around a…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-05-25","reactions":{}}