A new preview release of MPICH, 3.3b2, is now available for download. MPICH 3.3 contains a new (non-default) device layer implementation – CH4. CH4 is designed for low software overheads to better exploit next-generation hardware. An OFI (http://libfabric.org) or UCX (http://openucx.org) library is required to build CH4. Example configure lines: ./configure --with-device=ch4:ofi --with-libfabric=<path/to/ofi/install> ./configure --with-device=ch4:ucx --with-ucx=<path/to/ucx/install> CH4 is still in beta stages, meaning there are known build issues and bugs, but most tests and common benchmarks will complete on 64-bit Linux systems. Since 3.3b1, there have been more stability improvements, bug fixes, and code cleanup. PMIx client library support is added in CH4 to support launching with a compatible PMIx server. Also in this release is a reorganization of MPI collectives to make it easier to integrate new algorithms. A new framework is added to enable collective algorithms on either generic or device-specific functionality using a C++-template style system. Support is added for creating communicators based on hardware topology hints, and SLURM integration in Hydra is updated to work with the latest node list format. You can find the release on our downloads page (www.mpich.org/downloads). Regards, The MPICH team =============================================================================== Changes in 3.3 =============================================================================== # CH4 Device: A new device layer implementation designed for low software overheads. CH4 has experimental support for OFI and UCX network libraries, and POSIX shared memory. Thanks to Intel, Mellanox, and RIKEN AICS for participating in the CH4 coding effort. # Added support for splitting communicators based on hardware topology using info hints. # Fixed SLURM integration in Hydra for new node list format. # Added support for PMIx (https://pmix.github.io/pmix/) client library in CH4 netmods. Note that you must use a compatible PMIx server in this configuration. # Better organization of collectives in the MPI layer. The new scheme, which de-couples implementation from selection logic, enables easier integration of additional algorithms. # TSP collectives framework: A C++-template style framework for collective algorithms is added to allow single collective implementation to move data over generic or device-specific transport functions. # Improvements to derived datatype testing (DTPools - https://wiki.mpich.org/mpich/index.php/DTPools). # Added new "non-catastrophic" error codes to expose internal resource exhaustion. # Cleanup of whitespace (ch3 excluded) using the maint/code-cleanup.sh script. For instructions on how to update PRs/branches based on MPICH before the cleanup, see https://github.com/pmodels/mpich/wiki/Code-Cleanup-Procedure. # Removed the PAMI device and poe PMI client. # Several other minor bug fixes, memory leak fixes, and code cleanup. A full list of changes is available at the following link: http://git.mpich.org/mpich.git/shortlog/v3.2..v3.3b2 A list of bugs that have been fixed is available at the following link: https://github.com/pmodels/mpich/milestone/32?closed=1