Title :
A software architecture for global address space communication on clusters: put/get on fast messages
Author :
Giannini, Louis A. ; Chien, Andrew A.
Author_Institution :
Dept. of Comput. Sci., Illinois Univ., Urbana, IL, USA
Abstract :
Global address space parallel programming models can be an effective alternative to send/receive style communication, simplifying programming or code generation and increasing performance for certain application types. Traditionally, global address space mechanisms have been implemented in hardware in order to provide the necessary communication performance and responsiveness. However new high performance cluster messaging systems now allow global address space mechanisms to be realized efficiently in software. We describe a high performance one sided communication model that is implemented as a software layer on top of the Illinois Fast Messages (FM) system. We evaluate several different software implementation architectures for the remote agent, characterizing their differing performance characteristics. Our Put/Get FM implementation achieves peak bandwidths for put/get operations of 67 MBytes/s, overheads of a few microseconds, and remote read latencies as low as 26 microseconds on a Myrinet connected PC cluster. This implementation was released publicly as part of HPVM 1.0 in August 1997, and is receiving significant usage. It has been used for an implementation of the Global Arrays library and also serves as a back-end target for PGI´s commercial HPF compiler
Keywords :
message passing; parallel programming; storage allocation; 67 MByte/s; Global Arrays library; HPVM 1; Illinois Fast Messages system; Myrinet connected PC cluster; Put/Get FM implementation; back-end target; code generation; commercial HPF compiler; communication performance; global address space communication; global address space mechanisms; global address space parallel programming models; high performance cluster messaging systems; high performance one sided communication model; put/get on fast messages; put/get operations; remote agent; remote read latencies; software architecture; software implementation architectures; Application software; Bandwidth; Communication system software; Computer architecture; Delay; Hardware; Libraries; Parallel programming; Software architecture; Software performance;
Conference_Titel :
High Performance Distributed Computing, 1998. Proceedings. The Seventh International Symposium on
Conference_Location :
Chicago, IL
Print_ISBN :
0-8186-8579-4
DOI :
10.1109/HPDC.1998.710019