High performance multithreaded message passing on a myrinet cluster

The main purpose of this paper is to present the impact of pCoR thread level message passing facilities on applications running in a Myrinet cluster. To exploit Myrinet technology we use the GM library, which provides a limited number of ports as abstractions to name communication end-points. pCoR communication layer multiplexes GM ports by using a dispatcher thread to handle messages to/from a large number of communication entities (working threads). Our approach combines polling operations executed bymultiple threads to avoid unnecessary context switching; a simple mechanism is used to engage working threads into the polling scheme. We aim to reduce the total number of polling operations required to hold message passing system performance.

VIVO