[openib-general] Announcing the release of MVAPICH2 0.9.2 (MPI-2 for InfiniBand and other RDMA Interconnects)

[Dhabaleswar Panda]

The MVAPICH team is pleased to announce the release of MVAPICH2 0.9.2
with OpenIB/Gen2, VAPI and uDAPL transport interfaces. It also has
support for the standard TCP/IP (provided by MPICH2 stack). It is
optimized for the following platforms, OS, compilers and InfiniBand
adapters:

  - Platforms: EM64T, Opteron, IA-32 and Mac G5 
  - Operating Systems: Linux, Solaris and Mac OSX 
  - Compilers: gcc, intel, pathscale and pgi 
  - InfiniBand Adapters: Mellanox adapters with PCI-X 
    and PCI-Express (SDR and DDR with mem-full and mem-free cards) 

Starting with this release, MVAPICH2 enables InfiniBand support for
OpenIB/Gen2. All features available for the VAPI and uDAPL interfaces
are now available for the OpenIB/Gen2 interface.

MVAPICH2 0.9.2 is being distributed as a single integrated package
(with MPICH2 1.0.2p1 and MVICH).  It is available under BSD license.

This new release has the following features:

      - single code base with multiple underlying transport
        interfaces: VAPI, OpenIB/Gen2, uDAPL and TCP/IP
      - high-performance and optimized support for many MPI-2
        functionalities (one-sided, collectives, datatype)
      - support for other MPI-2 functionalities (as provided by 
        MPICH2 1.0.2p1). 
      - high-performance and optimized support for all MPI-1 
        functionalities (including two-sided)
      - high performance and optimized support for all one-sided 
        operations (Get, Put, and Accumulate)
      - support for both active and passive synchronization
      - optimized two-sided operations with RDMA support
      - efficient memory registration/de-registration schemes 
        for RDMA operations
      - optimized intra-node shared memory support (bus-based and
        NUMA)
      - shared library support for existing binary MPI programs to run
      - ROMIO support for MPI-IO
      - uDAPL support for portability across networks and OS 
        (tested for InfiniBand on Linux and Solaris; and Myrinet) 
      - scalable job start-up
      - optimized and tuned for the above platforms and different
        network interfaces (PCI-X and PCI-Express with SDR and DDR)
      - support for multiple compilers (gcc, icc, pathscale and pgi) 
      - single code base for all of the above platforms and OS
      - integrated and easy-to-use build script for installing the
        code on various platforms, OS, compilers, devices, and
        InfiniBand adapters
      - incorporates a set of runtime and compiler time tunable
        parameters for convenient tuning on large-scale clusters 

Other features of this release include:

- Excellent performance: Sample performance numbers include:
 
  - OpenIB/Gen2 on EM64T with PCI-Ex and IBA-DDR:
      Two-sided operations: 
        - 3.08 microsec one-way latency (4 bytes)
        - 1476 MB/sec unidirectional bandwidth 
        - 2661 MB/sec bidirectional bandwidth  

      One-sided operations:
        - 4.84 microsec Put latency 
        - 1483 MB/sec unidirectional Put bandwidth 
        - 2661 MB/sec bidirectional Put bandwidth 

  - OpenIB/Gen2 on EM64T with PCI-Ex and IBA-SDR:
      Two-sided operations:
        - 3.35 microsec one-way latency (4 bytes) 
        - 964 MB/sec unidirectional bandwidth
        - 1846 MB/sec bidirectional bandwidth 

      One-sided operations:
        - 5.43 microsec Put latency
        - 964 MB/sec unidirectional Put bandwidth
        - 1846 MB/sec bidirectional Put bandwidth

  - OpenIB/Gen2 on Opteron with PCI-Ex and IBA-SDR: 
      Two-sided operations:
        - 3.27 microsec one-way latency (4 bytes) 
        - 968 MB/sec unidirectional bandwidth
        - 1896 MB/sec bidirectional bandwidth 

      One-sided operations:
        - 5.95 microsec Put latency
        - 968 MB/sec unidirectional Put bandwidth
        - 1896 MB/sec bidirectional Put bandwidth

  - Solaris uDAPL/IBTL on Opteron with PCI-X and IBA-SDR:
      Two-sided operations:
        - 5.58 microsec one-way latency (4 bytes)
        - 655 MB/sec unidirectional bandwidth
        - 799 MB/sec bidirectional bandwidth

  - OpenIB/Gen2 uDAPL on Opteron with PCI-Ex and IBA-SDR:
      Two-sided operations:
        - 3.63 microsec one-way latency (4 bytes)
        - 962 MB/sec unidirectional bandwidth
        - 1869 MB/sec bidirectional bandwidth

  Performance numbers for all other platforms, system configurations
  and operations can be viewed by visiting `Performance Results'
  section of the project's web page.

- Similar performance with MVAPICH: With the new ADI-3-level design,
  MVAPICH2 0.9.2 delivers similar performance for two-sided operations
  compared to MVAPICH 0.9.6. Organizations and users interested in
  getting the best performance for both two-sided and one-sided 
  operations may migrate from MVAPICH code base to MVAPICH2 code base. 

- A set of benchmarks to evaluate both two-sided and one-sided
  operations (Put, Get, and Accumulate)

- An enhanced and detailed `User Guide' to assist users: 

       - to install this package on different platforms
            with interfaces (VAPI, uDAPL, OpenIB/Gen2 and TCP/IP) and
            different options

       - to vary different parameters of the MPI installation to 
            extract maximum performance and achieve scalability,
            especially on large-scale systems.

You are welcome to download the MVAPICH2 0.9.2 package and access
relevant information from the following URL:

http://nowlab.cse.ohio-state.edu/projects/mpi-iba/

A stripped down version of this release is available at the OpenIB SVN. 

A successive version with additional features and integrated with
MPICH2 1.0.3 will be available soon.

All feedbacks, including bug reports and hints for performance tuning,
are welcome. Please send an e-mail to mvapich-help at cse.ohio-state.edu.

Thanks, 

MVAPICH Team at OSU/NBCL 

----------

PS: If you would like to be removed from this mailing list, please end
an e-mail to mvapich_request at cse.ohio-state.edu.


======================================================================
MVAPICH/MVAPICH2 project is currently supported with funding from
U.S. National Science Foundation, U.S. DOE Office of Science,
Mellanox, Intel, Cisco Systems, Sun Microsystems and Linux Networx;
and with equipment support from AMD, Apple, Appro, IBM, Intel,
Mellanox, Microway, PathScale, SilverStorm and Sun Microsystems. Other
technology partner includes Etnus.
======================================================================