RECS - Resource Efficient Cluster System

The Project

In the project Resource Effienct Cluster System (RECS) we want to provide supercomputer performance to small and medium-sized enterprises. Typically, these companies do not have large budgets and adequate instructed personal to operate classical supercomputer systems.

Our Mission

Building affordable and simply administrable computer systems with low operating costs. The systems have to scale from small to large configurations to optimally fit application requirements.

The rapid development of the computer technology led to the fact that there is often more computational power available than needed for personal and small office applications. On the other hand, technical applications often need much more powerful computer systems than available. The development of new drugs, weather forecast, climate research, and physical simulations need even more powerful system, which can only be provided by supercomputers based on cluster systems. These clusters are collections of smaller computers connected by a high speed interconnect. Currently available clusters for high performance applications are often very expensive and customized.

The cooperation partners of this project develop a standardized und compact cluster system for high performance computing (HPC) and the high throughput computing (HTC) with the following design criteria:

  • high compute performance
  • low power consumption
  • low investment and administration costs
  • minimal required space

Altogether a resource efficient cluster server (RECS).

Partners

Funding

BMWi - Zentrales Innovationsprogramm Mittelstand (ZIM)

Contact

Research