Acceleration for MPI Derived Datatypes Using an Enhancer of Memory and Network

Author

Tanabe, Noboru ; Nakajo, Hironori

Author_Institution

Tokyo Univ. of Agric. & Technol., Tokyo, Japan

fYear

2008

fDate

21-23 Jan. 2008

Firstpage

17

Lastpage

23

Abstract

This paper presents a support function for MPI derived datatypes on DIMMnet-3 network interface with multi-banked extended memory, which is under development. Semi-hardwired derived datatype communication based on RDMA with hardwired gather and scatter is proposed. This mechanism and MPI using it are implemented and validated on DIMMnet-2 which is a former prototype operating on DDR DIMM slot. The performance of gather or scatter transfer of 8byte elements with large interval by using vector commands of DIMMnet-2 is 6.8 compared with software on a host. Proprietary benchmark of MPI derived datatype communication for transferring a submatrix corresponding to a narrow HALO area is executed. Observed bandwidth on DIMMnet-2 is far higher than that for similar condition with VAPI based MPI implementation on Infniband, even though poorer CPU and motherboard are used.

Keywords

message passing; network interfaces; DDR DIMM slot; DIMMnet-2; DIMMnet-3 network interface; HALO; MPI derived datatypes; RDMA; VAPI; datatype communication; memory enhancer; multibanked extended memory; network enhancer; Acceleration; Bandwidth; Field programmable gate arrays; Hardware; Network interfaces; Prototypes; Research and development; Scattering; Software performance; Supercomputers; MPI; RDMA; derived datatype; network interface;

fLanguage

English

Publisher

ieee

Conference_Titel

Innovative Architecture for Future Generation High-Performance Processors and Systems (IWIA), 2008 International Workshop on

Conference_Location

Hilo, HI

ISSN

1537-3223

Print_ISBN

978-1-4244-6465-4

Electronic_ISBN

1537-3223

Type

conf

DOI

10.1109/IWIA.2008.15

Filename

5453547