[mvapich-discuss] Announcing the release of MVAPICH2 2.2rc2, MVAPICH2-X 2.2rc2, and OMB 5.3.1
Panda, Dhabaleswar
panda at cse.ohio-state.edu
Mon Aug 8 17:29:19 EDT 2016
The MVAPICH team is pleased to announce the release of MVAPICH2
2.2rc2, MVAPICH2-X 2.2rc2 (Advanced MPI Features, Support for OSU INAM
and Hybrid MPI+PGAS (OpenSHMEM, UPC, CAF and UPC++) with Unified
Communication Runtime), and OSU Micro-Benchmarks (OMB) 5.3.1.
Features and enhancements for MVAPICH2 2.2rc2 are as follows:
* Features and Enhancements (since 2.2rc1):
- Enhanced performance for MPI_Comm_split through new bitonic
algorithm
- Thanks to Adam T. Moody at LLNL for the patch
- Enable graceful fallback to Shared Memory if LiMIC2 or CMA
transfer fails
- Enable support for multiple MPI initializations
- Unify process affinity support in Gen2, PSM and PSM2 channels
- Remove verbs dependency when building the PSM and PSM2 channels
- Allow processes to request MPI_THREAD_MULTIPLE when socket or
NUMA node level affinity is specified
- Point-to-point and collective performance optimization for Intel
Knights Landing
- Automatic detection and tuning for InfiniBand EDR HCAs
- Warn user to reconfigure library if rank type is not large
enough to represent all ranks in job
- Collective tuning for Opal at LLNL, Bridges at PSC, and
Stampede-1.5 at TACC
- Tuning and architecture detection for Intel Broadwell processors
- Add ability to avoid using --enable-new-dtags with ld
- Thanks to Adam T. Moody at LLNL for the suggestion
- Add LIBTVMPICH specific CFLAGS and LDFLAGS
- Thanks to Adam T. Moody at LLNL for the suggestion
* Bug Fixes (since 2.2rc1):
- Disable optimization that removes use of calloc in ptmalloc hook
detection code
- Thanks to Karl W. Schulz at Intel
- Fix weak alias typos (allows successful compilation with CLANG
compiler)
- Thanks to Min Dong at Old Dominion University for the patch
- Fix issues in PSM large message gather operations
- Thanks to Adam T. Moody at LLNL for the report
- Enhance error checking in collective tuning code
- Thanks to Jan Bierbaum at Technical University of Dresden for
the patch
- Fix issues with UD based communication in RoCE mode
- Fix issues with PMI2 support in singleton mode
- Fix default binding bug in hydra launcher
- Fix issues with Checkpoint Restart when launched with mpirun_rsh
- Fix Fortran binding issues with Intel 2016 compilers
- Fix issues with socket/NUMA node level binding
- Disable atomics when using Connect-IB with RDMA_CM
- Fix hang in MPI_Finalize when using hybrid channel
- Fix memory leaks
MVAPICH2-X 2.2rc2 provides support for advanced MPI features (User
Mode Memory Registration and Non-blocking Collectives with
Core-Direct), OSU INAM and hybrid MPI+PGAS (UPC, OpenSHMEM, CAF, and
UPC++) programming models with unified communication runtime for
emerging exascale systems. This library also provides flexibility for
users to write applications using the following programming models
with a unified communication runtime: MPI, MPI+OpenMP, pure UPC, pure
OpenSHMEM, pure UPC++, and pure CAF programs as well as hybrid
MPI(+OpenMP) + PGAS (UPC, OpenSHMEM, CAF, and UPC++) programs.
Features and enhancements for MVAPICH2-X 2.2rc2 are as follows:
* Features and Enhancements (since 2.2rc1):
- MPI Features
- Based on MVAPICH2 2.2rc2 (OFA-IB-CH3 interface)
- Efficient support for On Demand Paging (ODP) feature of
Mellanox for point-to-point and RMA operations
- Support for Intel Knights Landing architecture
- UPC Features
- Support for Intel Knights Landing architecture
- UPC++ Features
- Support for Intel Knights Landing architecture
- OpenSHMEM Features
- Support for Intel Knights Landing architecture
- CAF Features
- Support for Intel Knights Landing architecture
- Hybrid Program Features
- Support Intel Knights Landing architecture for hybrid
MPI+PGAS applications
- Unified Runtime Features
- Based on MVAPICH2 2.2rc2 (OFA-IB-CH3 interface). All the
runtime features enabled by default in OFA-IB-CH3 and
OFA-IB-RoCE interface of MVAPICH2 2.2rc2 are available in
MVAPICH2-X 2.2rc2
* Bug Fixes (since 2.2rc1):
- Fix compilation warnings and memory leaks
New features, enhancements and bug fixes for OSU Micro-Benchmarks
(OMB) 5.3.1 are listed here.
* New Features & Enhancements
- Add option to control whether CUDA kernels are built
- Add runtime option to specify number of threads for osu_latency_mt
* Bug Fixes
- Check if -lrt or -lpthread is needed
- Fix compilation warnings
- Fix non-blocking collective memory leak
- Correct documentation for osu_multi_lat
For downloading MVAPICH2 2.2rc2, MVAPICH2-X 2.2rc2, OMB 5.3.1 and
associated user guides, quick start guide, and accessing the SVN,
please visit the following URL:
http://mvapich.cse.ohio-state.edu
All questions, feedback, bug reports, hints for performance tuning,
patches and enhancements are welcome. Please post it to the
mvapich-discuss mailing list (mvapich-discuss at cse.ohio-state.edu).
Thanks,
The MVAPICH Team
More information about the mvapich-discuss
mailing list