Despite the variety of scientific, engineering and other areas in which complex systems need to be studied and modeled, the key information technology requirements for computational modeling and simulation of complex systems are essentially the same across many domains and applications. These requirements include:

1. Integration of large heterogeneous volumes of data and information that may arise from simulation or other information systems, which may be geographically widely dispersed.
2. Design and execution of compute- and memory-intensive simulation programs may require resources that are not available locally.
3. Handling the considerable volumes of data generated as output from the underlying simulations. These data need to be managed, analyzed and then shared using varied computational methodologies.

To address these requirements and achieve the main aims of the QosCosGrid project, the project is structured into the following key objectives:

First, to provide a quasi-opportunistic supercomputing grid architecture and infrastructure which includes

1. Necessary grid middleware services including monitoring and measurement capabilities,
2. User interfaces that enable easy access and use of resources by hiding the underlying complexity of the system,
3. Flexible fault-tolerant message passing libraries,
4. Data distribution enabling technology, and
5. Remote steering capabilities;

Second, to develop services that provide

1. Dynamic resource brokering giving the best quality-of-service to any given complex system simulation,
2. Reservation and orchestration of resources, communication, synchronization and routing as known from massively parallel processors computers.

Third to validate the quasi-opportunistic supercomputing concept with various types of complex systems simulation applications including

1. Research the non-trivial parallelization of the simulation- and data-processing applications typically encountered in the CS research, and to adapt the underlying algorithms to the quasi-opportunistic supercomputing environment,
2. Demonstration of the selected applications,
3. Provision of a testing environment.