Difference between revisions of "Bright:Using MPI"
Jump to navigation
Jump to search
(Created page with "The Message Passing Interface (MPI) is a standardized and portable message passing system, designed to function on a wide variety of parallel computers. The standard defines t...") |
|||
| Line 2: | Line 2: | ||
== Interconnects == | == Interconnects == | ||
| + | |||
| + | Jobs can use particular networks for intra-node communication: | ||
| + | |||
| + | *'''Gigabit Ethernet:''' Gigabit Ethernet is the interconnect that is most commonly available. For Gigabit Ethernet, no additional modules or libraries are needed. The Open MPI, MPICH implementations will work over Gigabit Ethernet. | ||
| + | *'''InfiniBand:''' InfiniBand is a high-performance switched fabric which is characterized by its high throughput and low latency. Open MPI, MVAPICH and MVAPICH2 are suitable MPI implementations for InfiniBand. | ||
Revision as of 14:41, 30 October 2014
The Message Passing Interface (MPI) is a standardized and portable message passing system, designed to function on a wide variety of parallel computers. The standard defines the syntax and semantics of a core of library routines useful to a wide range of users writing portable message-passing programs in Fortran or the C programming language. MPI libraries allow the compilation of code so that it can be used over a variety of multiprocessor systems from SMP nodes to NUMA systems and interconnected cluster nodes.
Interconnects
Jobs can use particular networks for intra-node communication:
- Gigabit Ethernet: Gigabit Ethernet is the interconnect that is most commonly available. For Gigabit Ethernet, no additional modules or libraries are needed. The Open MPI, MPICH implementations will work over Gigabit Ethernet.
- InfiniBand: InfiniBand is a high-performance switched fabric which is characterized by its high throughput and low latency. Open MPI, MVAPICH and MVAPICH2 are suitable MPI implementations for InfiniBand.