SOS12 Workshop
From petascale to exascale: Issues and hopes

Which science can first profit from peta/exascale machines? What has to be done to reach high efficiency, high availability, and high throughput on such highly parallel architectures? Is it hardware improvement, basic software development, better compilers, highly optimized libraries? Or must the application specialist make the major effort to get his application well running on thousands of many-core processing units? Which applications are already ready to go? Can Grid computing be an alternative to peta/exaflop machines? What are the solutions by the vendors, by the big research laboratories, by the universities, and by industry involved in applications demanding petaflop and beyond performance. These answers will be discussed during the SOS12 workshop. It is expected that the meeting continues to push co-operations between highly skilled research groups in ultrascale computation.

The programme concentrates on the software issues of the peta/exascale machines. Besides the basic software problems, two application domains are invited to present their solutions and expectations from the next generation machines.

Ultrascale science areas to be addressed include: nanoscience, biology, climate, chemistry, and others.