Difference between revisions of "Fortran implementation of the Actor Model using MPI"
From MohidWiki
| Line 1: | Line 1: | ||
A very simple ping-pong program is shown here. There are 2 programs, ''[[ping]]'' and ''[[pong]]'', that shoot messages among them, starting each with a random number of plays. The first program reaching 0 will send a message for the other program to stop and stops itself. File ''[[moduleMPImanagement]]'' is necessary to compile both programs. | A very simple ping-pong program is shown here. There are 2 programs, ''[[ping]]'' and ''[[pong]]'', that shoot messages among them, starting each with a random number of plays. The first program reaching 0 will send a message for the other program to stop and stops itself. File ''[[moduleMPImanagement]]'' is necessary to compile both programs. | ||
| + | If you examine the code there is no single MPI_BARRIER or any other explicit synchronization point. | ||
Revision as of 18:30, 30 November 2014
A very simple ping-pong program is shown here. There are 2 programs, ping and pong, that shoot messages among them, starting each with a random number of plays. The first program reaching 0 will send a message for the other program to stop and stops itself. File moduleMPImanagement is necessary to compile both programs. If you examine the code there is no single MPI_BARRIER or any other explicit synchronization point.
