Intel Cluster Tools

Intel MPI Library 4.0

Smart fabrics selection – Simplified usage through automatic choice of fastest transport protocol between MPI processes without additional environment settings. Fully configurable device and fallback selection by environment variables.
Enhanced dynamic connection establishment – Reduced memory footprint by introducing lazy mode connection, which establishes connections only when needed.
Two-phase communication buffer enlargement – Reduced memory consumption by allocating only the necessary communication buffer memory space.

DAPL intra-node communication mode – Bandwidth advantage for large messages by optional use of DAPL inside a multi-core or SMP node.
Further optimised collective operations – Significantly optimised versions of MPI_Reduce, MPI_Allreduce, MPI_Alltoall, MPI_Alltoallv, MPI_Broadcast, and others.
Enhanced process pinning – Maximised performance on Intel multi-core and SMP nodes. Provides for flexible process pinning.
Scalable job startup protocol – Application startup time significantly improved with this release.
Static version of libraries built without -fpic flag – Offers better performance for statically linked applications.

Supports a distributed install option that provides installation with Intel MPI Library 3.0 on the head node and compute nodes of a cluster, in one operation.

Additional thread safe libraries at level MPI_THREAD_MULTIPLE – An MPI application process can be multi-threaded and multiple threads may make MPI calls without restrictions.
Backward binary compatibility with Intel MPI Library 2.0 – Applications and objects compiled with Intel MPI Library 2.0 will work with the Intel MPI Library 3.0 run-time library.
Enhanced handling of multi-homed environment – Simplifies job and communication management and increases performance when working with multiple network interfaces per node.

Support for SLES 10
Enhanced Intel tool support
Intel® Trace Analyzer and Collector 7.0 – Enhanced analysis capability for system network activity when combined with Intel MPI Library 3.0, adding new features such as trace file comparison and performance counters.
Intel® Math Kernel Library Cluster Edition (Intel® MKL Cluster Edition) 9.0 – Get optimal performance when you combine Intel MKL Cluster Edition with Intel MPI Library.
Intel® MPI Benchmarks 3.0

Support for Intel® Debugger 8.1-23, 9.1-23 – Advancing productivity by running the Intel command line debugger on a parallel application using Intel MPI Library.
Etnus Totalview 6.8 and higher
Allinea DDT* 1.9.2 and higher

Rate this page

Comment