Wiss. Rechnen » System Information

Caution: this page contains information about the HoRUS cluster! The corresponding page for the OMNI cluster can be found here.

This page contains the technical details of the cluster hardware.


The cluster is divided into nodes. Each node has multiple cores and local memory (RAM). The hard drive storage is centralized. Access is via two login nodes (login1 and login2). The cluster occupies two water-cooled double racks inside the New Data Center on the Campus Hölderlinstraße.

Compute nodes

The cluster has 134 regular compute nodes. Each of these nodes has 12 CPU cores and 48 gigabytes of RAM. The nodes are named cn001-cn134. The CPUs are of type Intel Xeon X5650 (2 per node), with 6 cores per Xeon CPU. The CPU frequency is 2.66 GHz, the cache size is 12 MB.

Medium2 nodes

In addition to the regular compute nodes the cluster has 20 so-called fat nodes with more cores (16 instead of 12) and more memory (64 GB). These are available as the partition medium2. They were financed by the Lehrstuhl für Strömungstechnik und Strömungsmaschinen (Department of Mechanical Engineering), but are available to all cluster users at a reduced priority setting.

Other nodes

The cluster has one node (pre1) intended for interactive preprocessing. This node has 512 GB RAM and 32 cores.

Additionally there is a node named smp1 which also has 512 GB RAM and is available via the smp queue.


The cluster has a central file systen with a total of around 10 TB of storage space. This includes, among other things, the home directories of all users which are limited to 100 GB each. The Workspaces are on a separate but also centralized file system with a total size of around 60 TB. Individual workspaces are unlimited in size but have a limited duration of 30 days (can be extended three times by 30 days each time).


The nodes are connected via a fast interconnect (Infiniband) and are reachable (via Ethernet) from the internet.

Technical data:

  • Nodes:
    • Compute nodes: 134 x Intel Xeon X5650
      • CPU: 12 * 2.66 GHz, 12 MB Cache
      • RAM: 48 GB DDR3, 1333 MHz
    • Fat nodes: 20 * Intel Xeon Haswell EP E5-2630v3
      • CPU: 16 * 2.4 GHz, 20 MB Cache
      • RAM: 64 GB DDR4, 2133 MHz
    • pre1 and smp1 nodes: Intel Xeon E7-8837
      • 32 x 2,67 GHz, 24 MB Cache
      • RAM: 512 GB DDR3, 1066 MHz
  • Approximate total performance: ca. 17 TFlop/s
  • Storage:
    • Home directories: 8.5 Terabytes
    • Workspaces: 66 Terabytes
  • Interconnect:
    • Infiniband
    • Ethernet
  • Power consumption: ca. 30-35 kilowatts

Operating system

The operating system on the cluster is CentOS Linux Release 7.4.1708.

The cluster is being operated with the Bright Cluster Manager (Version 8.0).


Aktualisiert um 12:17 am 14. August 2018 von Jan Philipp Stephan