How to ... Select a Computer

NERSC has IBM POWER3 and POWER5 systems and a Linux cluster that provide High Performance Computing (HPC) platforms, and an analytics server that can be used to support calculations performed on the HPC systems. NERSC also hosts the Parallel Distributed Systems Facility (PDSF).

See the NERSC Systems page for more information.

Massively Parallel Processor (MPP) Systems

Franklin - Cray XT4

The NERSC Cray XT4 named franklin (franklin.nersc.gov), is a distributed memory machine with 9,669 compute nodes, each containing 2 CPU cores. Each processor is a 2.6 GHz Opteron. Nodes have from 4 GB of memory.

Bassi - IBM p575 POWER 5

The NERSC IBM POWER 5 system named bassi.nersc.gov, is a distributed memory machine with 111 compute nodes, each containing 8 processors. Each processor has a peak performance of 7.6 Gflops. Nodes have 32 GB of memory.

Jacquard - Opteron Linux Cluster

Jacquard is an Opteron cluster with 320 dual-processor nodes available to run scientific applications. The nodes are connected by a high-speed InfiniBand network and all have access to a large global GPFS file system.

Parallel Distributed Systems Facility

PDSF - Linux Cluster

The PDSF is a networked distributed computing environment used to meet the detector simulation and data analysis requirements of large scale High Energy Physics (HEP) and Nuclear Science (NS) investigations.

Analytics Server

DaVinci - SGI Altix

The NERSC Analytics Server, DaVinci, allows users to perform visualizations from remote locations. The machine is a Silicon Graphics Altix 350 with 32 1.4 GHz Itanium-2 processors, 192 GB of shared memory, and approximately 24 TB of disk space.

Mass Storage System

The NERSC mass storage system is the High Performance Storage System (HPSS).

Global File System