Silicon Graphics Altix XE1300, Quick Reference Manual

Page: 10 / 66

2 007-4979-004
1: SGI Altix XE1300 Cluster Quick-reference
systems have multiple processors (and/or multiple cores) that share memory, but the basic rule is
that a process is run on a dedicated processor core.
These are the primary hardware component types in the rackmounted cluster:
• Head node(s)
(SGI Altix XE240 or XE250 2U-high servers)
• Compute nodes
(SGI Altix XE310 or XE320 1U-high servers)
• Network interconnect components
(Gigabit Ethernet switches, InfiniBand switches, PCI cards, and cables)
The head node is connected to the interconnect network and also to the “outside world”, typically
via the local area network (LAN). The head node is the point of submittal for all MPI application
runs in the cluster. An MPI job is started from the head node and the sub-processes are distributed
to the cluster compute nodes from the head node. The main process on the head node will wait for
the sub-processes to finish. For large clusters or clusters that run many MPI jobs, multiple head
nodes may be used to distribute the load.
The compute nodes are identical computing systems that run the primary processes of MPI
applications. These compute nodes are connected to each other through the interconnect network.
The network interconnect components are typically Gigabit Ethernet or InfiniBand. The MPI
messages are passed across this network between the processes. This compute node network does
not connect directly to the “outside world” because mixing external and internal cluster network
traffic could impact application performance.
Site Plan Verification
Ensure that all site requirements are met before you install and boot your system. If you have
questions about the site requirements or you would like to order full-size floor templates for your
site, contact an SGI site planning representative by e-mail (site@sgi.com).
Additional helpful site planning information can be found in the SGI Altix XE Cluster Site
Planning Guide, (P/N 007-5456-001).