ACM TechNews, Friday, January 7, 2011
European Exascale Project Drives Toward Next Supercomputing Milestone
HPC Wire (01/06/11)

The goal of the European Exascale Software Initiative (EESI) is to
help effect the migration from petascale to exascale systems over
the next 10 years by bringing together industry and government
organizations. "The expected outputs of the project is an exascale
roadmap and set of recommendations to the funding agencies shared by
the European [high performance computing (HPC)] community, on
software--tools, methods, and applications--to be developed for this
new generation of supercomputers," says EESI program leader
Jean-Yves Berthou. EESI's first international workshop in Amsterdam
convened 80 experts in the fields of software development,
performance analysis, applications knowledge, funding models, and
governance aspects in HPC. Eight working groups (WGs)--four focused
on application grand challenges and four concentrating on enabling
exaflop computing methods--have been organized to identify and
classify the main challenges in their scientific area or technology
component. Netherlands National Computing Facilities foundation's
Peter Michielse says the workshops' purpose were twofold--to ensure
that each WG was mulling the correct challenges within its
scientific and technology discipline, and to become familiar with
Asian and U.S. initiatives with respect to their exascale software
projects. "An important role of EESI is to make sure that Europe is
involved in global discussions on hardware, software, and
applications design," Michielse says.

Better Benchmarking for Supercomputers
IEEE Spectrum (01/11) Mark Anderson
Many computer scientists say the High-Performance Linpack test used
to rate the world's Top 500 supercomputers is not the best
performance measurement for supercomputers. "What we're most
interested in is being able to traverse the whole memory of the
machine," says Sandia National Laboratory researcher Richard Murphy.
He and his colleagues have developed the Graph500, a new benchmark
that which rates supercomputers based on gigateps (billions of
traversed edges) instead of petaflops. By the Graph500 standard,
supercomputers have actually been slowing down, according to Notre
Dame University professor Peter Kogge. Over the past 15 years, each
1,000-fold increase in flops has resulted in a 10-fold decrease in
accessible memory. According to the Graph500 standard, the top
supercomputer would be Argonne National Laboratory's IBM Blue
Gene-based Intrepid, which recorded 6.6 gigateps. The U.S. Defense
Advanced Research Projects Agency, the Department of Energy, and the
National Science Foundation also have developed a new benchmark
called the HPC Challenge, which tests computing power and memory
accessibility.