Performance prediction and procurement in practice: assessing the suitability of commodity cluster components for wavefront codes

The cost of state-of-the-art supercomputing resources makes each individual purchase an expensive and, in many cases, lengthy process. Often each candidate architecture will need to be benchmarked using a variety of tools to assess potential performance. However, benchmarking alone often provides only limited insight into the potential scalability and suitability of each architecture for procurement. In this paper we present a case study applying two recently developed performance models to the Chimaera benchmarking code written by the United Kingdom Atomic Weapons Establishment (AWE) with a view to analysing how the code will perform and scale on a medium sized, commodity based InfiniBand cluster. Our models are validated with average accuracies of 90% against an existing InfiniBand machine and then used as the basis for predicting code performance on a variety of hardware configurations including changes in the underlying network, faster processors and high core density per processor. The results of our experimentation with machine performance parameters demonstrate the compute-bound nature of Chimaera and its sensitivity to network latency at increased processor counts. By using these insights we are able to discuss potential strategies which may be employed during the procurement of future mid-range clusters for a wavefront-code rich workload.