The distributed virtual shared-memory system based on the InfiniBand architecture

Inho Park, Seon Wook Kim

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Even though there have been strong research activities about distributed virtual shared-memory (DVSM) systems, their architectures have been not widely used in current high-performance computing markets. The reason is that the previously introduced DVSM systems use conventional interconnection technologies like Ethernet, which incurs high execution overhead due to process interruption at data communication for memory consistency. In this paper, we present the DVSM architecture based on the next generation of an interconnection technique, the InfiniBand Architecture (IBA). Because the IBA supports shared-memory programming semantics by means of remote direct-memory access (RDMA) and atomic operations in hardware, we can minimize the communication overhead for memory consistency on the DVSM system. For characterizing multithreaded applications on our IBA-based DVSM system, we examined two different shared-memory programming models, i.e. SPMD and OpenMP benchmarks. We show that our DVSM to use full features of the IBA can improve the performance significantly over the IPoIB-based DVSM system in all benchmarks, and also comparable to the bus-based shared-memory multiprocessor system in some benchmarks.

Original languageEnglish
Pages (from-to)1271-1280
Number of pages10
JournalJournal of Parallel and Distributed Computing
Volume65
Issue number10
DOIs
Publication statusPublished - 2005 Oct 1

Fingerprint

Virtual Memory
InfiniBand
Shared Memory
Data storage equipment
Benchmark
Interconnection
Computer programming
Architecture
Shared-memory multiprocessors
Computer systems
Memory Model
Data Communication
Ethernet
OpenMP
Multiprocessor Systems
System Architecture
Memory architecture
Programming Model
Communication
Programming

Keywords

  • Distributed virtual shared-memory
  • InfiniBand
  • OpenMP
  • Performance
  • SPMD

ASJC Scopus subject areas

  • Computer Science Applications
  • Hardware and Architecture
  • Control and Systems Engineering

Cite this

The distributed virtual shared-memory system based on the InfiniBand architecture. / Park, Inho; Kim, Seon Wook.

In: Journal of Parallel and Distributed Computing, Vol. 65, No. 10, 01.10.2005, p. 1271-1280.

Research output: Contribution to journalArticle

@article{7bd90e4ccdec42908b14bcd1b2b4bca2,
title = "The distributed virtual shared-memory system based on the InfiniBand architecture",
abstract = "Even though there have been strong research activities about distributed virtual shared-memory (DVSM) systems, their architectures have been not widely used in current high-performance computing markets. The reason is that the previously introduced DVSM systems use conventional interconnection technologies like Ethernet, which incurs high execution overhead due to process interruption at data communication for memory consistency. In this paper, we present the DVSM architecture based on the next generation of an interconnection technique, the InfiniBand Architecture (IBA). Because the IBA supports shared-memory programming semantics by means of remote direct-memory access (RDMA) and atomic operations in hardware, we can minimize the communication overhead for memory consistency on the DVSM system. For characterizing multithreaded applications on our IBA-based DVSM system, we examined two different shared-memory programming models, i.e. SPMD and OpenMP benchmarks. We show that our DVSM to use full features of the IBA can improve the performance significantly over the IPoIB-based DVSM system in all benchmarks, and also comparable to the bus-based shared-memory multiprocessor system in some benchmarks.",
keywords = "Distributed virtual shared-memory, InfiniBand, OpenMP, Performance, SPMD",
author = "Inho Park and Kim, {Seon Wook}",
year = "2005",
month = "10",
day = "1",
doi = "10.1016/j.jpdc.2005.04.008",
language = "English",
volume = "65",
pages = "1271--1280",
journal = "Journal of Parallel and Distributed Computing",
issn = "0743-7315",
publisher = "Academic Press Inc.",
number = "10",

}

TY - JOUR

T1 - The distributed virtual shared-memory system based on the InfiniBand architecture

AU - Park, Inho

AU - Kim, Seon Wook

PY - 2005/10/1

Y1 - 2005/10/1

N2 - Even though there have been strong research activities about distributed virtual shared-memory (DVSM) systems, their architectures have been not widely used in current high-performance computing markets. The reason is that the previously introduced DVSM systems use conventional interconnection technologies like Ethernet, which incurs high execution overhead due to process interruption at data communication for memory consistency. In this paper, we present the DVSM architecture based on the next generation of an interconnection technique, the InfiniBand Architecture (IBA). Because the IBA supports shared-memory programming semantics by means of remote direct-memory access (RDMA) and atomic operations in hardware, we can minimize the communication overhead for memory consistency on the DVSM system. For characterizing multithreaded applications on our IBA-based DVSM system, we examined two different shared-memory programming models, i.e. SPMD and OpenMP benchmarks. We show that our DVSM to use full features of the IBA can improve the performance significantly over the IPoIB-based DVSM system in all benchmarks, and also comparable to the bus-based shared-memory multiprocessor system in some benchmarks.

AB - Even though there have been strong research activities about distributed virtual shared-memory (DVSM) systems, their architectures have been not widely used in current high-performance computing markets. The reason is that the previously introduced DVSM systems use conventional interconnection technologies like Ethernet, which incurs high execution overhead due to process interruption at data communication for memory consistency. In this paper, we present the DVSM architecture based on the next generation of an interconnection technique, the InfiniBand Architecture (IBA). Because the IBA supports shared-memory programming semantics by means of remote direct-memory access (RDMA) and atomic operations in hardware, we can minimize the communication overhead for memory consistency on the DVSM system. For characterizing multithreaded applications on our IBA-based DVSM system, we examined two different shared-memory programming models, i.e. SPMD and OpenMP benchmarks. We show that our DVSM to use full features of the IBA can improve the performance significantly over the IPoIB-based DVSM system in all benchmarks, and also comparable to the bus-based shared-memory multiprocessor system in some benchmarks.

KW - Distributed virtual shared-memory

KW - InfiniBand

KW - OpenMP

KW - Performance

KW - SPMD

UR - http://www.scopus.com/inward/record.url?scp=24144443844&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=24144443844&partnerID=8YFLogxK

U2 - 10.1016/j.jpdc.2005.04.008

DO - 10.1016/j.jpdc.2005.04.008

M3 - Article

VL - 65

SP - 1271

EP - 1280

JO - Journal of Parallel and Distributed Computing

JF - Journal of Parallel and Distributed Computing

SN - 0743-7315

IS - 10

ER -