Automatic Task-Based Code Generation for High Performance Domain Specific Embedded Language

Providing high level tools for parallel programming while sustaining a high level of performance has been a challenge that techniques like Domain Specific Embedded Languages try to solve. In previous works, we investigated the design of such a DSEL —NT 2 —providing a Matlab  -like syntax for paralle...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International journal of parallel programming Jg. 44; H. 3; S. 449 - 465
Hauptverfasser: Tran Tan, Antoine, Falcou, Joel, Etiemble, Daniel, Kaiser, Hartmut
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York Springer US 01.06.2016
Springer Nature B.V
Schlagworte:
Algorithms
Analysis
Automation
C plus plus
Computer programming
Computer Science
Cost control
Design analysis
Design engineering
Language
Level (quantity)
Libraries
Linear algebra
Matlab
New technology
Parallel processing
Parallel programming
Processor Architectures
Programming languages
Run time (computers)
Semantics
Software Engineering/Programming and Operating Systems
Studies
Theory of Computation
Parallel skeletons
C++
Generative programming
Asynchronous programming
ISSN:0885-7458, 1573-7640
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Providing high level tools for parallel programming while sustaining a high level of performance has been a challenge that techniques like Domain Specific Embedded Languages try to solve. In previous works, we investigated the design of such a DSEL —NT 2 —providing a Matlab  -like syntax for parallel numerical computations inside a C++  library. In this paper, we show how NT 2  has been redesigned for shared memory systems in an extensible and portable way. The new NT 2  design relies on a tiered Parallel Skeleton system built using asynchronous task management and automatic compile-time taskification of user level code. We describe how this system can operate various shared memory runtimes and evaluate the design by using two benchmarks implementing linear algebra algorithms.
Bibliographie:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ObjectType-Article-1
ObjectType-Feature-2
content type line 23
ISSN:0885-7458
1573-7640
DOI:10.1007/s10766-015-0354-9