ICS 2014

28th International Conference on Supercomputing


One fullday and four half day workshops are scheduled for ICS 2014. They will be held on Tuesday, June 10th.

W1: ROSS 2014 - Runtime and Operating Systems for Supercomputers (Full day)

The workshop’s objective is to gather the researchers from academia, national labs, and industry, from all over the world, working in the area of operating system software and runtime systems for supercomputers and, during the course of the workshop, to facilitate the exchange of ideas, information, and experiences.

W2: International Workshop on Auto-Tuning of HPC Applications (Half day)

The tuning of applications for high-performance computing systems has fundamentally been linked with advances in parallel performance tools. The techniques to measure and analyze performance behavior in order to understand bottlenecks and remove limiting factors has evolved hand-in-hand with each generation of HPC technologies. A new era of supercomputing is upon us where traditional analysis methods that involve post-mortem and human-in-the-loop approaches will be challenged by more complex performance interactions, dynamic runtime execution, and multi-objective optimizations. A promising strategy is to apply auto-tuning methodologies to enhance the performance analysis environment and add sophisticated support for optimization search.

The AutoTune consortium and Super project would like to bring together people interested in tools for tuning HPC applications. The idea is to present the results of the two projects and to put it into context with other projects working into the same direction. This workshop is continuing the very successful Dagstuhl workshop held in 2013 about automatic application tuning.

W3: Power and Performance Efficiencies for HPC (Half day)

With Exascale systems on the horizon, we have ushered in an era with power and energy consumption as the primary concerns for scalable computing. To achieve a viable Exaflop high performance computing capability, revolutionary methods are required with a stronger integration among hardware features, system software and applications. Equally important are the capabilities for fine-grained spatial and temporal measurement and control to facilitate these layers for energy efficient computing across all layers. This workshop seeks to address the important energy efficiency aspects in the HPC community. Emphasis is given to the application's view related to significant energy efficiency improvements and to the required hardware/software stack that must include necessary power and performance measurement and analysis harnesses. Uses include design-space exploration for future systems as well as dynamic optimization for current systems. The workshop aims to bring together researchers from different communities working on challenging problems in this area for a dynamic exchange of ideas.

On a different but important note our strategy for putting together the workshop will now be centered around invited experts in order to put together the highest quality program that should be of interest to many attending ICS. We will still be open to people who would like to make position statements on topics of interest.

W4: Extreme Scaling (Half day)

In June 2014 LRZ is going to organize an on-site four day workshop on extreme code scaling which will enable scaling of user implemented software on SuperMUC, the 3 PetaFLOP/s machine at LRZ.

LRZ Users can submit their own software as candidates for the workshop. Max. 6-8 projects will finally be selected for runs on the whole available system during the 4-day workshop.

At the ICS2014, the participants of the on-site workshop will present their codes and report about the results obtained during the on-site workshop. The talks will also cover topics like lessons learned, challenges, i/o subsystems, and parallelisation techniques.

Deadline for proposal submission to the on-site LRZ workshop is: April 1st, 2014.

W5: Eploiting Different Levels of Parallelism for Exascale Computing (Half day)

Today, many scientific computing applications feature different levels of parallelism induced, e.g., by the increasing complexity of models, advanced solver methods, or parallelization in time.

Often, an open question is how to exploit this in the actual implementation on peta- and exascale architectures. The workshop is organized by the coordinators of the projects EXAHD and ExFSA in the priority program 1648, Software for Exascale Computing, of the German Research Foundation, contributions from other projects of this priority program are particularly welcome.