Scalable Interpolation on GPUs for Thermal Fluids Applications
- Argonne National Lab. (ANL), Argonne, IL (United States); Univ. of Tennessee, Knoxville, TN (United States)
- Argonne National Lab. (ANL), Argonne, IL (United States); Univ. of Illinois at Urbana-Champaign, IL (United States)
- Argonne National Lab. (ANL), Argonne, IL (United States)
The need for efficient interpolation arises in many use cases for spectral element simulations. Towards this end, we present a fast and robust GPU-enable interpolation utility for NekRS, a spectral-element, Navier-Stokes solver. This utility is based on the findpts and findpts eval routines from the gslib library. To demonstrate the functionality, we added support for particle tracking. The GPU implementation of the interpolation allows for tracking 100 times more particles for the same computational cost.
- Research Organization:
- Argonne National Laboratory (ANL), Argonne, IL (United States)
- Sponsoring Organization:
- USDOE Office of Science. Exascale Computing Project
- DOE Contract Number:
- AC02-06CH11357
- OSTI ID:
- 1823469
- Report Number(s):
- ANL-21/55; 171648
- Country of Publication:
- United States
- Language:
- English
Similar Records
NekRS, a GPU-accelerated spectral element Navier–Stokes solver
High-order matrix-free incompressible flow solvers with GPU acceleration and low-order refined preconditioners
Energy Exascale Computational Fluid Dynamics Simulations With the Spectral Element Method
Journal Article
·
Mon Oct 17 20:00:00 EDT 2022
· Parallel Computing
·
OSTI ID:2429555
High-order matrix-free incompressible flow solvers with GPU acceleration and low-order refined preconditioners
Journal Article
·
Sun Apr 19 20:00:00 EDT 2020
· Computers and Fluids
·
OSTI ID:1735799
Energy Exascale Computational Fluid Dynamics Simulations With the Spectral Element Method
Journal Article
·
Wed Feb 21 19:00:00 EST 2024
· Journal of Fluids Engineering
·
OSTI ID:2572333