Research at the department for Parallel and Distributed Systems focuses on all (hardware-near) aspects to improve the execution performance of complex computational tasks. This covers multiple aspects, from smarter code and data placement to better resource control:

Smart Execution Environments

The PDS department is part of the MyThOS development group, which focuses on designing and developing a distributed Operating System to support flexible task execution on any connected device in a heterogeneous infrastructure. The OS ensures connectivity and maintenance of state independent of location and connectivity. The OS is designed in a fashion that no single instance becomes the bottleneck, but instead so that it allows for distribution of essential OS services over the distributed infrastructure. By pursuing a modular architecture, any OS instance can focus on selected services relevant for the task and device, without having to offer the full scope of OS functionalities. Implicitly, MyThOS reduces the adaptation effort for heterogeneous devices. 

Current research topics include:

• Live migration
• Elastic task and data placement
• P2P infrastructure sniffing

More information here soon.

Efficient Computing

More information here soon

Smart Global Data Analysis and Integration

More information here soon

Non-Deterministic Problem Solving

More information here soon