Lapedo: Hybrid skeletons for programming heterogeneous multicore machines in Erlang

Vladimir Janjic, Christopher Brown, Kevin Hammond

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

We describe Lapedo, a novel library of hybrid parallel skeletons for programming heterogeneous multi-core/many-core CPU/GPU systems in Erlang. Lapedo's skeletons comprise a mixture of CPU and GPU components, allowing skeletons to be flexibly and dynamically mapped to available resources, with all of the low-level tedious code to divide work between CPUs and GPUs, transfer the data between the main and GPU memory and offload computations to the GPUs provided by the library. We evaluate the effectiveness of Lapedo on three realistic use cases from different domains, demonstrating significant improvements in speedups compared to CPU-only and GPU-only executions.

Original languageEnglish
Title of host publicationParallel Computing
Subtitle of host publicationOn the Road to Exascale
EditorsFrans Peters, Mark Parsons, Mark Sawyer, Hugh Leather, Gerhard R. Joubert
PublisherIOS Press
Pages185-195
Number of pages11
Volume27
ISBN (Electronic)9781614996217
ISBN (Print)9781614996200
DOIs
Publication statusPublished - 2016

Publication series

NameAdvances in Parallel Computing
Volume27
ISSN (Print)0927-5452
ISSN (Electronic)1879-808X

Keywords

  • GPU offloading
  • Heterogeneous multicore systems
  • Hybrid skeletons
  • Parallel skeletons

ASJC Scopus subject areas

  • General Computer Science

Fingerprint

Dive into the research topics of 'Lapedo: Hybrid skeletons for programming heterogeneous multicore machines in Erlang'. Together they form a unique fingerprint.

Cite this