DMCA
Encapsulating synchronization and load-balance in heterogeneous programming (2012)
| Venue: | SpringerVerlag LNCS Series |
| Citations: | 3 - 1 self |
BibTeX
@INPROCEEDINGS{Torres12encapsulatingsynchronization,
author = {Yuri Torres and Arturo Gonzalez-escribano and Diego Llanos},
title = {Encapsulating synchronization and load-balance in heterogeneous programming},
booktitle = {SpringerVerlag LNCS Series},
year = {2012},
publisher = {Hitmap}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Abstract. Programming models and techniques to exploit parallelism in accelerators, such as GPUs, are different from those used in traditional parallel models for shared- or distributed-memory systems. It is a challenge to blend different programming models to coordinate and exploit devices with very different characteristics and computation powers. This paper presents a new extensible framework model to encapsulate runtime decisions related to data partition, granularity, load balance, synchronization, and communication for systems including assorted GPUs. Thus, the main parallel code becomes independent of them, using internal topology and system information to transparently adapt the computation to the system. The programmer can develop specific functions for each architecture, or use existent specialized library functions for different CPU-core or GPU architectures. The high-level coordination is expressed using a programming model built on top of message-passing, providing portability across distributed- or shared-memory systems. We show with an example how to produce a parallel code that can be used to efficiently run on systems ranging from a Beowulf cluster to a machine with mixed GPUs. Our experimental results show how the run-time system, guided by hints about the computational-power ratios of different devices, can automatically part and distribute large computations across heterogeneous systems, improving the overall performance. 1
Keyphrases
heterogeneous programming different characteristic parallel code experimental result computational-power ratio overall performance library function new extensible framework model load balance beowulf cluster different cpu-core gpu architecture traditional parallel model computation power large computation system information programming model run-time system distributed-memory system assorted gpus data partition main parallel code runtime decision mixed gpus specific function different device high-level coordination internal topology heterogeneous system shared-memory system
