Stanford Linear Accelerator Center
Advanced Computations
About Us
People
Research
Projects
Collaborations
Codes
Computers
Presentations
Publications
Reports
Support
Jobs
Links
Contact
Home
ACD has access to the world's largest unclassified supercomputer, NERSC IBM SP which has 380 compute nodes with 16 Power3 375 MHz processors per node. Its computing power is number 5 in the world according to June 2003's TOP 500 list. For more information visit http://hpcf.nersc.gov.

In addition, we also have a 500G storage server and another on the way. We also have a Linux cluster with 16 nodes connected by Myrinet, we are expecting delivery of 64-node Myrinet Linux cluster.

IBM SP - Seaborg --
The NERSC IBM SP RS/6000, named Seaborg, is a distributed memory machine with 6,080 compute processors. Each processor has a peak performance of 1.5 GFlops.

The processors are distributed among 380 compute nodes with 16 processors per node. Each compute node has between 16 and 64 GBytes of memory (4 have 64 GBytes; 64 have 32 GBytes; 312 have 16 GBytes).

The compute nodes contain processors and memory connected to each other with a high-bandwidth, low-latency switching network. Each node runs its own full instance of the standard AIX operating system. The disk storage system is a distributed, parallel I/O system called GPFS. Some nodes serve exclusively as GPFS servers.

Each node of Seaborg is a POWER3 SMP wide node which consists of 16 POWER3 processors and shared memory of 16 to 64 GBytes per node. Currently, 64 nodes have 32 GBytes of memory, and 4 nodes have 64 GBytes of memory.

For more information on High Performance Computing (HPC) click here.

2575 Sand Hill Rd.
MS 27
Menlo Park, CA 94025

Tel: 650.926.2864
Fax: 650.926.4603

acd@slac.stanford.edu