mirror of
https://github.com/nwchemgit/nwchem.git
synced 2026-07-26 21:25:29 -04:00
51 lines
1.3 KiB
HTML
51 lines
1.3 KiB
HTML
<HTML>
|
|
<!-- DO NOT EDIT: this is a cleared public file -->
|
|
|
|
<HEAD><TITLE>Supported Platforms </TITLE></HEAD>
|
|
<BODY>
|
|
|
|
<H1>Platforms on which PeIGS runs</H1>
|
|
|
|
<P>
|
|
PeIGS is designed for parallel and distributed memory computers with a
|
|
relatively high startup time for communication. PeIGS can also be run on
|
|
networks of workstations and shared memory computers using
|
|
Robert Harrison's
|
|
<A HREF="/docs/global/tcgmsg.html"> TCGMSG </A>
|
|
message passing software or using MPI.
|
|
|
|
<P>
|
|
PeIGS uses only blocking send & receive which are implemented in a portable
|
|
way. Switching to different communications libraries, e.g., PVM,
|
|
requires modifying only a few routines.
|
|
|
|
<HR>
|
|
|
|
PeIGS runs using the native message passing library on:
|
|
<UL>
|
|
<LI>Intel Delta, Paragon
|
|
</UL>
|
|
|
|
PeIGS currently runs using
|
|
<A HREF="/docs/global/tcgmsg.html"> TCGMSG </A> on:
|
|
|
|
<UL>
|
|
<LI>IBM SP series
|
|
<LI>Cray T3D, T3E
|
|
<LI>KSR-2
|
|
<LI>SGI Power Challenge, Origins, Octanes
|
|
<LI>Network and cluster of workstations(Unix-like, linux)
|
|
</UL>
|
|
|
|
PeIGS runs using MPI on:
|
|
<UL>
|
|
<LI>Intel Delta, Paragon
|
|
<LI>IBM SP series
|
|
<LI>SGI Power Challenge, Origins, Octanes
|
|
<LI>Network of workstations(Unix-like)
|
|
</UL>
|
|
|
|
PeIGS can be run using MPI on other machines by making a simple change
|
|
to the machine specific parts of a single Makefile.
|
|
</BODY>
|
|
</HTML>
|