POP maintains a co-design data repository which includes statistics about common performance issues of HPC applications as well as micro-kernels extracted from real applications each characterising fundamental performance behaviour. It is available at the website
Hardware architects or system software designers from other EU projects will be able to get quantitative information of how to estimate the potential impact of an architectural or system software approach they may be developing. The micro-kernels will also be useful within POP (but also outside) as training material examples, and for demonstrating benefits of the programming model features and practices that POP promotes in dissemination activities
The co-design data repository can be browsed in various ways:
- One angle are the performance metrics for different applications developed by and used within the POP Center of Excelence. These metrics allow characterizing the behavior of the applications and could be queried by system designers (architecture, system software) in projects outside POP to demonstrate the potential of their proposed approaches and get a rough estimate of which codes their techniques will have an important impact on.
- We also describe typical behavioural patterns (that may result in potential performance degradation) that we have identified in the analysis of applications using different algorithms in different disciplines.
- The objective then is to identify such patterns in generic terms, provide links to programs that expose them and links to the correspondent best-practice(s) that should address it. Some of these programs include additional versions that allow comparing the behaviour before/after applying a given best-practice.
- Programs included in this site use different programming languages and leverage some of the most popular parallel programming models used in the HPC community.
- Finally, reports from the performance assessments of the applications are available.