The system consists of a total of 102 computer nodes. These are divided into different departments.
- 9 compute nodes are available for theoretical chemistry, each with four AMD Opteron 6274 (16 cores per processor) with a clock speed of 2.2GHz and 32MB cache per chip. Furthermore, they have 128GB (7 knots) respectively 256GB (2 knots) RAM and 2TB local memory.
- 24 compute nodes are available for hydrogeology. These are each equipped with two Intel Xeon L5530 (8 cores per processor), with a clock frequency of 2.4GHz and 8MB cache per chip. Each node has 72GB RAM and 256GB of local SSD storage.
- 15 compute nodes are available for computational physics, each with two AMD Opteron 6172 (12 cores per processor), with a clock frequency of 2.1GHz and 18MB cache per chip. In addition, each node has 32GB RAM and 160GB of local storage.
- 6 compute nodes are available to the Massive Star Formation, each with two Intel Xeon E5-2670 v3 (24 cores per processor) with a clock speed of 2.3GHz and 30MB cache. Furthermore, each node has 128GB RAM and 600GB of local memory.
- 23 compute nodes are available for Earth System Dynamics, each with two Intel Xeon E5-2660 v3 (10 cores per processor) with a clock speed of 2.6GHz and 25MB cache. Furthermore, each node has between 128 and 256GB RAM and locally either 4TB SAS, 1TB SAS or 240GB SSD memory.
- 7 compute nodes are available for quantum optics, each with two Intel Xeon E5-2650 v3 (10 cores per processor) with a clock speed of 2.3GHz and 25MB cache. Each node has 128GB RAM and 2TB local SAS memory.
- 15 compute nodes are available to the Uni Group, each with two Intel Xeon 5150 (2 cores per processor) with a clock speed of 2.66GHz and 4MB cache. In addition, each node has 32GB RAM and 160GB (Raid 1) memory.
- 2 compute nodes are available for quantum optics and theoretical physics, each with two Intel Xeon E7-4870 (20 cores per processor) with a clock frequency of 2.4GHz and 30MB cache. There are also 768GB RAM and 12.6TB local memory (2 nodes) respectively 1TB RAM and 4TB (Raid 5) local memory (1 node).
There is a high-speed network via InfiniBand. As a global workspace file system, the scalable, parallel file system BeeGFS is used, which provides a total of 76TB (gross) temporary usable storage space for current calculations.
