Design and Implementation of a High Performance MPI for Large Scale Computing System

Author

Prabu, D. ; Vanamala, V. ; Deka, Sanjeeb Kumar ; Sridharan, R. ; Rao, B.B.P. ; Mohanram, N.

Author_Institution

Centre for Dev. of Adv. Comput., Bangalore

fYear

2007

fDate

2-4 April 2007

Firstpage

929

Lastpage

932

Abstract

The necessity to process large application in reasonable time has lead to the use of parallel processing and the use of programming standards, which enable the running of user application in a parallel environment. This paper presents C-DAC message passing interface (C-MPI). C-MPI is an optimized MPI for cluster of multiprocessors (CLUMPS). It has been able to exhibit higher performance compared to public domain MPI such as MPICH. In this work we have presented the design flow of C-MPI and compared the various performance of C-MPI and MPICH over PARAMNet-II interconnect using KSHIPRA as the underlying protocol

Keywords

message passing; multiprocessor interconnection networks; parallel processing; protocols; standards; large scale computing system; message passing interface; multiprocessors cluster; parallel processing; programming standards; protocol; Bandwidth; Communication system control; Concurrent computing; High performance computing; Large-scale systems; Message passing; Parallel programming; Protocols; Standards development; TCPIP;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Technology, 2007. ITNG '07. Fourth International Conference on

Conference_Location

Las Vegas, NV

Print_ISBN

0-7695-2776-0

Type

conf

DOI

10.1109/ITNG.2007.66

Filename

4151811