Gallery of Projects Using MPICH
This page provides a sampling of projects that are using MPICH, the high-performance,
portable implementation of the MPI
The MPICH implementation has been designed to provide an implementation of MPI
that is both high-performance and easy to port to new environments. If you
have used MPICH in your work, we'd like to hear from you. Project titles, web
pages, and papers are desired. Send any information to Bill Gropp. Thanks!
This page is currently incomplete.
is a open source parallel software for multi
dimensional data aggregation (OLAP). Multi dimensional
expression (MDX) query syntax is supported. Its highly
scalable and can run on network of workstations.
Research into MPI implementations
- MPICH-V is a fault-tolerant MPI
implementation, based on
MPICH. It relies on
stable channel memory servers and checkpoint servers. The checkpointing is
completely automated, thus the user do not need to modify his original code
in order to use the full system.
is a high performance
implementation of MPI on the emerging InfiniBand interconnect
standard. High performance implementations of both point-to-point and
collective communication operations are emphasized. This implementation was
developed at Ohio State.
MyVIA a high-performance cluster implementation of the Message Passing
Interface (MPI) based on a port of MVICH to VIA.
It's now running on Linux2.2.x SMP and Linux2.4.x SMP using Myrinet Lanai4.x
or Lanai9.x. The source code is available.
MPICH-PM/CLUMP is a
implementation of MPI using Myrinet.
Our MPICH implementation, now called MPICH-PM/CLUMP, is now running on
Linux 2.0.XX and Linux 2.1.1XX using Myrinet.
Our system called SCore cluster system software is a total software
architecture in a sense that it includes not only the communication
library MPI but also a global operating system, multi-threaded C++
language system, its debugger, and several utilities. Everyting runs
on top of Linux, NetBSD, and SunOS.
The global operating system called SCore-D enables a multi-users
environment without sacrificing the communication performance.
We employ the gang scheduling method to share the CPU and network
The SCore system runs on Intel Pentiums and Compaq Alphas.
- Zero-copy message transfer, and
- SMP cluster implementation
GAMMA Project: Genoa Active Message MAchine. A port of MPICH ontop of GAMMA
stands for Multi-Platform MPICH. It is a modification and extension
to the MPICH distribution (currently release 1.4). MP-MPICH compiles and
runs on all common UNIX platforms (just like the original MPICH) and also on
Windows NT (tested with Visual C++ 6.0).
The download link is: www.lfbs.rwth-aachen.de/content/mp-mpich-dl.
is an implementation of MPI using MPICH for SCI (Scalable Coherent
Interconnect) connected clusters, and is now a part of MP-MPICH.
MPI-FM is a
high-performance cluster implementation of the Message Passing Interface (MPI)
based on a port of MPICH to Fast Messages. The
Message Passing Interface is an
industry standard communication interface for message-passing parallel
programs. It provides a wealth of capabilities including synchronous and
asynchronous messaging, datatypes, and communicators. MPI-FM is a complete
implementation of the MPI standard 1.0 based on the Argonne/MSU MPICH code
base. However, the MPICH code base was tuned significantly to avoid buffer
copies and reduce the critical path length for message reception. The
effective software overhead for the MPI send/receive is below 3 microseconds
MPICH-GM is a version of MPICH
on top of Myricom's GM.
MPI-BIP is a port of
MPICH over Myrinet network using the BIP communication layer.
It is available for evaluation.
System Software Architecture for High-End Computing describes the software
for the Intel TFLOPS system at Sandia, including the MPI implementation that
is based on MPICH.
is a high-performance version of MPICH from the Argonne National Laboratory
and Mississippi State University, that has been extensively optimized for
Compaq's Alpha-based SMP systems and MEMORY CHANNEL clusters.
Compaq MPI uses a lightweight message-passing scheme to communicate over
shared memory and MEMORY CHANNEL, to dramatically reduce software overhead and
provide low latency, high bandwidth communications.
Other Tools and Packages That Include MPICH
OSCAR (Open Source Cluster Application Resources) is a bundle of
software designed to make it easy to build, maintain, and use a modest
sized Linux cluster.
Bibliography of Papers