2019/12/01 by Aditya Kashi, Siva Nadarajah, Kashi, Aditya +1
Computer Science · Engineering · Physics and Astronomy · #65F08 #65N22 #65Y05 #Advanced Numerical Methods in Computational Mathematics #Distributed #Electromagnetic Scattering and Analysis #FOS: Computer and information sciences #FOS: Mathematics #Matrix Theory and Algorithms #Numerical Analysis (math.NA) #Parallel #and Cluster Computing (cs.DC)
paper · pdf · doi:10.48550/arxiv.1912.00539
openalex publication_date 2019/12/01 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
We present a study of the effectiveness of asynchronous incomplete LU\nfactorization preconditioners for the time-implicit solution of compressible\nflow problems while exploiting thread-parallelism within a compute node. A\nblock variant of the asynchronous fine-grain parallel preconditioner adapted to\na finite volume discretization of the compressible Navier-Stokes equations on\nunstructured grids is presented, and convergence theory is extended to the new\nvariant. Experimental (numerical) results on the performance of these\npreconditioners on inviscid and viscous laminar two-dimensional steady-state\ntest cases are reported. It is found, for these compressible flow problems,\nthat the block variant performs much better in terms of convergence, parallel\nscalability and reliability than the original scalar asynchronous ILU\npreconditioner. For viscous flow, it is found that the ordering of unknowns may\ndetermine the success or failure of asynchronous block-ILU preconditioning, and\nan ordering of grid cells suitable for solving viscous problems is presented.\n