Journal Article FZJ-2022-03907

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png
A Deterministic Portable Parallel Pseudo-Random Number Generator for Pattern-Based Programming of Heterogeneous Parallel Systems

 ;  ;  ;

2022
Springer Science + Business Media B.V. Dordrecht [u.a.]

International journal of parallel programming 50(3-4), 319 - 340 () [10.1007/s10766-022-00726-5]

This record in other databases:    

Please use a persistent id in citations:   doi:

Abstract: SkePU is a pattern-based high-level programming model for transparent program execution on heterogeneous parallel computing systems. A key feature of SkePU is that, in general, the selection of the execution platform for a skeleton-based function call need not be determined statically. On single-node systems, SkePU can select among CPU, multithreaded CPU, single or multi-GPU execution. Many scientific applications use pseudo-random number generators (PRNGs) as part of the computation. In the interest of correctness and debugging, deterministic parallel execution is a desirable property, which however requires a deterministically parallelized pseudo-random number generator. We present the API and implementation of a deterministic, portable parallel PRNG extension to SkePU that is scalable by design and exhibits the same behavior regardless where and with how many resources it is executed. We evaluate it with four probabilistic applications and show that the PRNG enables scalability on both multi-core CPU and GPU resources, and hence supports the universal portability of SkePU code even in the presence of PRNG calls, while source code complexity is reduced.

Classification:

Contributing Institute(s):
  1. Jülich Supercomputing Center (JSC)
Research Program(s):
  1. 5122 - Future Computing & Big Data Systems (POF4-512) (POF4-512)
  2. EXA2PRO - Enhancing Programmability and boosting Performance Portability for Exascale Computing Systems (801015) (801015)

Appears in the scientific report 2022
Database coverage:
Medline ; Creative Commons Attribution CC BY 4.0 ; OpenAccess ; Clarivate Analytics Master Journal List ; Current Contents - Engineering, Computing and Technology ; DEAL Springer ; Ebsco Academic Search ; Essential Science Indicators ; NationallizenzNationallizenz ; SCOPUS ; Science Citation Index Expanded ; Web of Science Core Collection
Click to display QR Code for this record

The record appears in these collections:
Document types > Articles > Journal Article
Workflow collections > Public records
Institute Collections > JSC
Publications database
Open Access

 Record created 2022-10-27, last modified 2023-01-23


OpenAccess:
Download fulltext PDF
External link:
Download fulltextFulltext by OpenAccess repository
Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)