Skip to Main content Skip to Navigation
Reports

Do moldable applications perform better on failure-prone HPC platforms?

Abstract : This paper compares the performance of di erent approaches to tolerate failures using checkpoint/restart when executed on large-scale failure-prone platforms. We study (i) Rigid applications, which use a constant number of processors throughout execution; (ii) Moldable applications, which can use a di erent number of processors after each restart following a fail-stop error; and (iii) GridShaped applications, which are moldable applications restricted to use rectangular processor grids (such as many dense linear algebra kernels). For each application type, we compute the optimal number of failures to tolerate before relinquishing the current allocation and waiting until a new resource can be allocated, and we determine the optimal yield that can be achieved. We instantiate our performance model with realistic applicative scenarios and make it publicly available for further usage.
Document type :
Reports
Complete list of metadatas

Cited literature [18 references]  Display  Hide  Download

https://hal.inria.fr/hal-01799498
Contributor : Equipe Roma <>
Submitted on : Thursday, May 24, 2018 - 6:11:49 PM
Last modification on : Friday, January 10, 2020 - 3:10:18 PM
Long-term archiving on: : Saturday, August 25, 2018 - 2:52:47 PM

File

rr9174.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01799498, version 1

Collections

Citation

Valentin Le Fèvre, George Bosilca, Aurelien Bouteiller, Thomas Hérault, Atsushi Hori, et al.. Do moldable applications perform better on failure-prone HPC platforms?. [Research Report] RR-9174, Inria Grenoble Rhône-Alpes. 2018, pp.1-24. ⟨hal-01799498⟩

Share

Metrics

Record views

173

Files downloads

295