The main goal of this project is to enable PCRaster models to use the full power of available hardware, including CPUs and GPUs. With the advent of heterogeneous hardware platforms we have the opportunity to re-implement the PCRaster building blocks and to make all PCRaster models run faster. This will require re-implementing the individual PCRaster operations to let them use multiple CPU cores and GPUs. This work will improve all PCRaster models, because the core of PCRaster will be improved. By speeding up the core PCRaster build blocks, all existing and future PCRaster models will perform significantly better. A huge advantage is that existing models do not have to be updated for this.
A second goal of this project is to work with a number of PCRaster users at research institutes and companies to verify that the updated software indeed speeds up their existing models significantly, and to investigate possible new modelling opportunities with them that were previously not feasible.