Distributed NVRAM Cache – Optimization and Evaluation with Power of Adjacency Matrix - Computer Information Systems and Industrial Management (CISIM 2017) Access content directly
Conference Papers Year : 2017

Distributed NVRAM Cache – Optimization and Evaluation with Power of Adjacency Matrix

Artur Malinowski
  • Function : Author
  • PersonId : 1024469
Paweł Czarnul
  • Function : Author
  • PersonId : 1024470

Abstract

In this paper we build on our previously proposed MPI I/O NVRAM distributed cache for high performance computing. In each cluster node it incorporates NVRAMs which are used as an intermediate cache layer between an application and a file for fast read/write operations supported through wrappers of MPI I/O functions. In this paper we propose optimizations of the solution including handling of write requests with a synchronous mode, additional modes preventing data preloading from a file and synchronization on file close if the solution is used as temporary cache only. Furthermore, we have evaluated the solution for a real application that computes powers of an adjacency matrix of a graph in parallel. We demonstrated superiority of our solution compared to a regular MPI I/O implementation for various powers and numbers of graph nodes. Finally, we presented good scalability of the solution for more than 600 processes running on a large HPC cluster.
Fichier principal
Vignette du fichier
448933_1_En_2_Chapter.pdf (589.03 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01656233 , version 1 (05-12-2017)

Licence

Attribution

Identifiers

Cite

Artur Malinowski, Paweł Czarnul. Distributed NVRAM Cache – Optimization and Evaluation with Power of Adjacency Matrix. 16th IFIP International Conference on Computer Information Systems and Industrial Management (CISIM), Jun 2017, Bialystok, Poland. pp.15-26, ⟨10.1007/978-3-319-59105-6_2⟩. ⟨hal-01656233⟩
267 View
137 Download

Altmetric

Share

Gmail Facebook X LinkedIn More