Welcome to this web page related to "MPI over InfiniBand, 10GigE/iWARP and RDMA over Converged Ethernet (RoCE)" project, lead by Network-Based Computing Laboratory (NBCL) of The Ohio State University. The MVAPICH2 software, supporting MPI 3.0 standard, delivers best performance, scalability and fault tolerance for high-end computing systems and servers using InfiniBand, 10GigE/iWARP and RoCE networking technologies. This software is being used by more than organizations world-wide in 75 countries (Current Users) to extract the potential of these emerging networking technologies for modern systems. As of , more than downloads have taken place from this project's site. This software is also being distributed by many InfiniBand, 10GigE/iWARP and RoCE vendors in their software distributions. The MVAPICH2-X software package provides support for hybrid MPI+PGAS (UPC and OpenSHMEM) programming models with unified communication runtime for emerging exascale systems. The MVAPICH2-GDR package provides support for clusters with NVIDIA GPUs supporting the GPUDirect RDMA feature. The MVAPICH2-MIC package provides support for clusters with Intel MIC coprocessors. MVAPICH2 software is powering several supercomputers in the TOP 500 list. Examples (from the November '14 ranking) include:
- 7th, 519,640-core (Stampede) at TACC
- 11th, 160,768-core (Pleiades) at NASA
- 15th, 76,032-core (Tsubame 2.5) at Tokyo Institute of Technology
This project is supported by funding from U.S. National Science Foundation, U.S. DOE Office of Science, Ohio Board of Regents, ODOD, Cisco Systems, Cray, Intel, Linux Networx, Mellanox, NVIDIA, QLogic, and Sun Microsystems; and equipment donations from Advanced Clustering, AMD, Appro, Chelsio, Dell, Fulcrum, Fujitsu, Intel, Mellanox, Microway, NetEffect, QLogic and Sun. Other technology partner includes: TotalView Technologies.
(NEW) MVAPICH2-X 2.1rc2 providing support for hybrid MPI+PGAS (UPC, OpenSHMEM and CAF) programming models, support for UH CAF 3.0.9 with efficient point-to-point read/write, CO_REDUCE and CO_BROADCAST collectives, support for OpenSHMEM 1.0h, and support for UPC 2.20.0. [more]
(NEW) MVAPICH2 2.1rc2 (based on MPICH 3.1.4) with enhanced startup performance, CR support with DMTCP, large message RMA, optimized collectives for 4K process, and enhancements for PSM interface. [more]
Three Presentations on MPI, PGAS and Multicasting with GPUDirect RDMA at the upcoming NVIDIA GTC 2015 conference.
MVAPICH2-GDR 2.1a (based on MVAPICH2 2.1a) with optimized small message transfer, R3 support for GPU based packetized transfer, support for MPI_Scan and MPI_Exscan from GPU buffers and optimized collectives support using GPU Direct RDMA (GDR) for NVIDIA GPUs. [more]
MVAPICH2-MIC 2.0 (based on MVAPICH2 2.0.1) with optimized pt-to-pt and collective support for native, symmetric and offload modes on clusters with Intel MICs (Xeon Phis) is available. [more]