Towards Scalable Unstructured Mesh Computations on Shared Memory Many-Cores
Open Access
- 20 February 2024
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 109-119
Abstract
Due to data conflicts or data dependences, exploiting shared memory parallelism on unstructured mesh applications is highly challenging. The prior approaches are neither general nor scalable on emerging many-core processors. This paper presents a general and scalable shared memory approach for unstructured mesh computations. We recursively divide and reorder an unstructured mesh to construct a task dependency tree (TDT), where massive parallelism is exposed and data conflicts as well as data dependences are respected. We propose two recursion strategies to support popular programming models on both CPUs and GPUs for TDT. We evaluate our approach by applying it to an industrial unstructured Computational Fluid Dynamics (CFD) software. Experimental results show that our approach significantly outperforms the prior shared memory approaches, delivering up to 8.1× performance improvement over the engineer-tuned implementations.Keywords
Funding Information
- National Natural Science Foundation of China (62272474)
- Ministry of Science and Technology of China (2021YFB0300101)
This publication has 23 references indexed in Scilit:
- Parallelization of an unstructured Navier–Stokes solver using a multi-color ordering method for OpenMPComputers & Fluids, 2013
- Some useful strategies for unstructured edge‐based solvers on shared memory machinesInternational Journal for Numerical Methods in Engineering, 2010
- A Fast and High Quality Multilevel Scheme for Partitioning Irregular GraphsSIAM Journal on Scientific Computing, 1998
- An assessment of one- and two-equation turbulence models for internal and external flows1997
- Renumbering unstructured grids to improve the performance of codes on hierarchical memory machinesAdvances in Engineering Software, 1997
- CilkACM SIGPLAN Notices, 1995
- SOLVING SPARSE TRIANGULAR LINEAR SYSTEMS ON PARALLEL COMPUTERSInternational Journal of High Speed Computing, 1989
- Lower-upper Symmetric-Gauss-Seidel method for the Euler and Navier-Stokes equationsAIAA Journal, 1988
- Approximate Riemann solvers, parameter vectors, and difference schemesJournal of Computational Physics, 1981
- Nested Dissection of a Regular Finite Element MeshSIAM Journal on Numerical Analysis, 1973