Why we are developing HPCW:
The quality and the accuracy of climate and weather forecasts are limited by the computing power available. In order to measure and evaluate the computationa performance of climate and weather models and the suitability of different computer architectures for these models, we need benchmarks. Classical benchmarks, like the High Performance Linpack (HPL) on which the TOP500 list is based are not representative of typical model codes used for climate research and weather prediction.
