Criticism of the European Commission's supercomputing research policy: it is too much centralized, says Peter Welchering on Deutschlandfunk

29 Jun 2016 Frankfurt - Peter Welchering says on Deutschlandfunk that he noted a growing criticism of the EU HPC policy at the ISC conference in Frankfurt. The 1,2 billion euro Human Brain Project for instance, largely a supercomputing project, does not deliver on its promises and is now turning towards classical neural networking techniques. Thes types of large projects are not good for application diversity which was one of the strengths of Europe, Welchering says. He also heard a lot of criticism on the fact that every country has to get its fair share of HPC capacity, even if that means just keeping the existing systems alive and do not do anything new. Welchering also heard criticism on the strong focus on exascale computing. There should be a much broader view on spending research funding. The podcast of the broadcast (in German) is still available.
Ad Emmen