Interactive, data-intensive HPC

<!--HTML--><div> <div> <div> <p>Complex scientific questions at large scale user facilities strongly&nbsp;involve the user at every step of experimental planning. The data deluge&nbsp;these users have to deal with is strongly coupled to an increasing&nbsp;difficulty in inverting data unambiguously. Thus, combining inverse&nbsp;methods with forward solutions during experiments will become the norm,&nbsp;making, challenging existing data analysis workflows that aim for&nbsp;automatic real time data reduction. In this talk we discuss both&nbsp;challenges and paths to address them, fusing high performance computing&nbsp;with desktop-like, interactive workflows that enable users&nbsp;inexperienced in HPC to use large-scale data analysis effectively.</p> </div> </div> </div>