D8.1.2: Performance Model of Community Codes

This document presents the results achieved by PRACE-2IP Work Package 8 at PM3. Scientific communities, selected during the first project month, proposed a number of codes relevant for the scientific domain and promising in terms of potential performance improvement on the coming generation of supercomputing architectures. In order to analyse the performance features of these codes, a methodology, based on the performance modelling approach, has been defined and adopted. This methodology relies on the analytic modelling of the main algorithms (characterising the dependency from the critical model parameters) and on the performance analysis, based on the usage of performance tools. The performance modelling approach allows studying the current behaviour of a code, emphasising performance and bottlenecks. But it is also a predictive tool, allowing the estimation of the code’s behaviour on different computing architectures, and identifying the most promising areas for performance improvement. The first step of the modelling is represented by performance analysis. This analysis was accomplished for all the proposed codes, with detailed data generated and collected. The overall results are presented for each scientific domain and code. Most of the data was generated for “real cases”, i.e., running the codes for real scientifically meaningful cases, in order to evaluate performances and bottlenecks in daily usage configurations, and to achieve a performance impact through code refactoring and optimisation of the crucial sections of each code. In the following step of WP8, this data will be synthesised and modelled, in order to identify the most promising numerical kernels for performance improvement. This will be the subject of the coming deliverable D8.1.3 “Prototype Codes Exploring Performance Improvements”