Performance models and workload distribution algorithms for optimizing a hybrid CPU-GPU multifrontal solver

DOI10.1016/j.camwa.2014.01.013zbMath1416.65125OpenAlexW2022648602MaRDI QIDQ316653

Chenhan D. Yu, Wei-Chung Wang

Publication date: 27 September 2016

Published in: Computers \& Mathematics with Applications (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/j.camwa.2014.01.013

zbMATH Keywords

multifrontal methodsparse linear systemsymmetric positive definitegraphics processing unit (GPU)model based performance optimization

Mathematics Subject Classification ID

Computational methods for sparse matrices (65F50) Parallel numerical computation (65Y05) Direct numerical methods for linear systems and matrix inversion (65F05) Numerical algorithms for specific classes of architectures (65Y10)

Uses Software

• LAPACK
• BLAS
• MUMPS
• CUBLAS
• MKL
• StarPU
• DAGuE

Cites Work

• A Fully Asynchronous Multifrontal Solver Using Distributed Dynamic Scheduling
• Multifrontal parallel distributed symmetric and unsymmetric solvers
• Multifrontal Computations on GPUs and Their Multi-core Hosts
• A Scalable High Performant Cholesky Factorization for Multicore with GPU Accelerators
• The Use of Linear Graphs in Gauss Elimination
• The Multifrontal Solution of Indefinite Sparse Symmetric Linear
• The Multifrontal Method for Sparse Matrix Solution: Theory and Practice
• The influence of relaxed supernode partitions on the multifrontal method
• A set of level 3 basic linear algebra subprograms
• Implementing Multifrontal Sparse Solvers for Multicore Architectures with Sequential Task Flow Runtime Systems
• A column pre-ordering strategy for the unsymmetric-pattern multifrontal method

This page was built for publication: Performance models and workload distribution algorithms for optimizing a hybrid CPU-GPU multifrontal solver