Computational Cluster Programs

Hoffman2 Cluster Hardware

Shared Cluster Hardware System Standards for Compute Nodes

In order to maximize the effective management of the Shared Cluster System and to provide the highest level computing services to Shared Cluster customers, compute nodes added to the Hoffman2 Cluster must meet minimum standards. As of Auguest 2010 these standards are:

  • 1U, rack mounted compute nodes. Half-width "dual" nodes preferred.
  • High quality, well integrated, hardware in terms of thermal, power, rack mounting capability and parts used
  • Dual-six-core 2.6GHz Intel Nehalem 2.66GHz CPUs
  • 4GB of memory per core
  • 160-250GB hard drive per node
  • Gigabit Ethernet port
  • DDR/QDR InfiniBand interconnect
  • PCI-Express slot (required for high-speed interconnect) if not on motherboard.
  • 3 year warranty
  • The preferred minium node order is two. If you cannot purchase two nodes please contact Bill Labate, labate@ats.ucla.edu, X67323, to discuss possible alternatives.

These standards will be evaluated periodically and updated based on best price/performance.

Campus General Purpose Cluster Nodes

The Campus General Purpose Cluster consists of 256 cores on nodes configured to the same specifications as the rest of the Shared Cluster. This is done to allow, by prior special agreement, the use of the entire cluster by large compute runs.

Network and Interconnect

The Hoffman2 Cluster has both an InfiniBand interconnect and a gigabit Ethernet network. The Ethernet network is dedicated to traffic in and out of the storage system and also handles various administrative functions. InfiniBand is used for inter-node, MPI-type communication. Using these two interconnects maximizes performance within the cluster.

Ethernet Network

Multiple Cisco 6509-E 288 port GigE switches with a redundant 10Gb uplink to the campus backbone.

InfiniBand

Multiple Cisco and Q-Logic Infiniband switches.

Storage

BlueArc

A high-performance, fault-tolerant BlueArc storage system provides storage for the Hoffman2 Cluster System. The BlueArc disk space is formatted as RAID 5 (striped with distributed parity) to increase performance. See BlueArc Storage System for additional hardware details.

The BlueArc's very low latency results in high data throughput. It can hold up to 8 billion total files with up to 4 million files per directory.

On the Hoffman2 Cluster home directories are served from the BlueArc. Users of the General Purpose Cluster Nodes currently have a 20GB quota on their home directories. Users of the Shared Cluster System pay a one-time $1,500 per terabyte, fee for storage. This fee pays for storage space for 3 years as well as administration of the storage system and backup services for a mutually agreed to amount of data.

Panasas Storage System

See Panasas Storage System: redundancy and usable space for additional hardware details.