Robust Memory-Aware Mappings for Parallel Multifrontal Factorizations

Abstract : We study the memory scalability of the parallel multifrontal factorization of sparse matrices. We illustrate why commonly used mapping strategies (e.g. proportional mapping) cannot achieve a high memory efficiency. We propose a class of "memory-aware' algorithms that aim at maximizing performance under memory constraints. These algorithms provide both accurate memory predictions and a robust solver. We illustrate our approach with experiments performed on large matrices with the MUMPS solver.
Complete list of metadatas

https://hal.inria.fr/hal-00726644
Contributor : Emmanuel Agullo <>
Submitted on : Thursday, August 30, 2012 - 9:45:33 PM
Last modification on : Tuesday, October 29, 2019 - 7:36:06 AM

Identifiers

  • HAL Id : hal-00726644, version 1

Citation

François-Henry Rouet, Emmanuel Agullo, Patrick R. Amestoy, Alfredo Buttari, Abdou Guermouche, et al.. Robust Memory-Aware Mappings for Parallel Multifrontal Factorizations. SIAM Conference on Parallel Processing for Scientific Computing (SIAM PP 2012), Feb 2012, Savannah, Georgia, United States. ⟨hal-00726644⟩

Share

Metrics

Record views

564