Computational power of HPC systems continues to grow, while the cost of individual computers continues to decline. Distributed computing techniques allow thousands of processors to participate in a single simulation. While this results in simulations that are more accurate and useful for predicting complex phenomena, it also requires large amounts of data to be saved and analyzed to glean knowledge from the results of a simulation.

To avoid data bottlenecks, Kitware offers a co-processing toolset in ParaView that integrates core data processing with simulations to enable scalable data analysis. The co-processing toolset has two main parts: an extensible and flexible co-processing library and configuration tools for co-processing configuration.