CUMULVS is a software infrastructure for the collaborative execution of large, multi-component distributed simulations. It supports interactive visualization and remote steering of distributed applications by multiple collaborators, and provides fault tolerance to applications running in heterogeneous distributed environments.
CUMULVS: Collaborative Infrastructure for Computational Steering, Interactive Visualization, and Fault Tolerance in Distributed Environments
In light of the growing emphasis towards computer simulation and remote control of Virtual Laboratory experiments, CUMULVS could impact the entire scientific community by empowering collaborating scientists to visually monitor large data fields of an ongoing simulation and to remotely control parameters inside a running application.
Computational steering has the potential to revolutionize computational experiments by allowing scientists to interactively explore (steer) a simulation in time and/or space, and concentrate more on the science than on the computer. Through the use of on-the-fly interaction, the computer will become a more useful tool to the engineer, allowing experimentation and real time exploration of a design space. CUMULVS takes this one step further by allowing a distributed group of experts to view and steer a large scale simulation in order to accelerate the cycle of scientific discovery.
The CUMULVS system has been integrated into several DOE Computational Grand Challenges as well as DOE 2000 applications. Shown below is such an example where a combustion computation based on POET (Parallel Object-oriented Environment and Toolkit) has been distributed across the supercomputers at several DOE laboratories.
Scientists at each of these sites as well as remote collaborators use CUMULVS to visualize the turbulent combustion process during the computation and on-the-fly adjust parameters related to the chemically reacting flow. CUMULVS also supplies the failure detection and recovery required to run large scale applications over NGI.
The development of CUMULVS is funded by DOE's Mathematical, Information, and Computational Sciences Division.