Solving complex models with high
accuracy and within a reasonable computing time motivates
the development of domain decomposition methods in space,
time or other variables that make an efficient use of modern
high performance computing (HPC) systems. The current
top HPC architectures have already attained million-way
concurrency, and current trends suggest that processor
counts will continue to grow rapidly. Both from the
pertinence of the approach and the challenges it raises
stands the decomposition of the time domain because the
sequential nature of the evolution makes it difficult to
reconcile with efficient parallelization methods.
The aim of this workshop, the 7th
of the series, is to bring together researchers working on
parallel-in-time methods to discuss recent advances and
future directions. It will take place in the marine
station
of Roscoff (France), which offers the
unique combination of a quiet and friendly environment for
academic meetings and immediate closeness to the city and
its wonderful marine and coastal surroundings.