Skip to Main content Skip to Navigation
Conference papers

Controlling the Memory Subscription of Distributed Applications with a Task-Based Runtime System

Abstract : This talk discuss the need for cooperation between a task-based distributed application code and a run-time system engine to control the memory subscription levels throughout the execution. We show that the task paradigm allows to control the memory footprint of the application by throttling the task submission flow rate, striking a compromise between the performance benefits of anticipative task submission and the resulting memory consumption. We illustrate the benefits of our contribution on a compressed dense linear algebra distributed application.
Complete list of metadata

https://hal.inria.fr/hal-01380126
Contributor : Marc Sergent <>
Submitted on : Wednesday, October 12, 2016 - 3:04:07 PM
Last modification on : Sunday, September 20, 2020 - 6:25:44 PM
Long-term archiving on: : Saturday, February 4, 2017 - 8:02:59 PM

File

siampp.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01380126, version 1

Collections

Citation

Marc Sergent, David Goudin, Samuel Thibault, Olivier Aumage. Controlling the Memory Subscription of Distributed Applications with a Task-Based Runtime System. SIAM Conference on Parallel Processing for Scientific Computing (SIAM PP 2016), Apr 2016, Paris, France. pp.318 - 327. ⟨hal-01380126⟩

Share

Metrics

Record views

475

Files downloads

210