Big data challenges arising from future experiments

Future physics experiments and observatories rely on the capabilitiesto process significantly larger data streams. Examples are futurelight-source experiments, elementary particle experiments andradio-astronomy observatories. All have in common that they planto exploit high-performance compute capabilities instead of relyingon hardware controlled data processing. This approach increasesflexibility during the lifetime of such experiments and may increasethe use of commodity hardware, which is typically cheaper compared tocustom solutions. While these experiments can thus benefit from HPCarchitectures and technologies, both as being available today as wellas planned on future roadmaps, the requirements and use models differsignificantly from today's high-performance computing. In this talkwe will analyse the requirements of a few examples and discuss how theywill benefit from current as well as future HPC technologies