G8 (2011 - 2014)
Enabling Climate Simulation at Extreme Scale
Policy decisions for mitigating climate change or adapting to it are subjects of great discussion throughout the world. Uninformed decisions will impose a heavy cost on future generations, both financial and human. Therefore, it is essential to reduce the current uncertainties about future climate changes and their impact by running climate simulations at 1,000 times larger scales than today. Exascale supercomputers (1018 operations per second) will appear in 2018-2020, featuring a hierarchical design and gathering 100 millions of computing cores. The numerical models of the physics, chemistry, and biology affecting the climate system need to be improved to run efficiently on these extreme systems. Without improvement, these codes will not produce simulations results required to respond to the societal and economical challenges of climate change.
The objective of the G8 ECS (Enabling Climate Simulation at Extreme Scale) project is to investigate how to run efficiently climate simulations on future Exascale systems and get correct results. This project gathers top researchers in climate and computer science to focus on three main topics: (i) how to complete simulations with correct results despite frequent system failures, (ii) how to exploit hierarchical computers with hardware accelerators close to their peak performance and (iii) how to run efficient simulations with 1 billion threads. This project also aims at educating new generations of climate and computer scientists about techniques for high performance computing at extreme scale.
The consortium, lead by Marc Snir (director) from the University of Illinois at Urbana Champaign and Franck Cappello (associate director) from the French National Institute for Research in Computer Science and Control, gathers researchers from Canada (University of Victoria), France (INRIA), Germany (German Research School for Simulation Sciences), Japan (Tokyo Tech and University of Tsukuba), Spain (Barcelona Supercomputing Center) and USA (University of Illinois at Urbana Champaign, University of Tennessee, National Center for Supercomputing Applications, and National Center for Atmospheric Research). It will use the top supercomputers to experiment with new techniques in the previously described three topics. The team of Prof. Wolf will investigate the performance of climate codes using the Scalasca performance-analysis tool, which his group jointly develops together with the group of Dr. Bernd Mohr at the Jülich Supercomputing Centre. Experimental studies on Jugene, Jülich's IBM Blue Gene/P leadership system, will be supported by Dr. Lars Hoffmann from the local Simulation Laboratory for Climate Modeling.
This three-year project receives G8 coordinated funding from the Natural Sciences and Engineering Research Council of Canada (NSERC), French National Research Agency (ANR), German Research Foundation (DFG), Japan Society for the Promotion of Science (JSPS) and National Science Foundation (NSF). This project, together with five other projects, was funded as part of the G8 Research Councils Initiative on Multilateral Research, Interdisciplinary Program on Application Software towards Exascale Computing for Global Scale Issues. This is the first initiative of its kind to foster broad international collaboration on the research needed to enable effective use of future exascale platforms.
More information can be found in HPCwire, an online magazine for high performance computing.
G8 Project Partners
University of Illinois at Urbana-Champaign, USA
Department of Computer Science
University of Tennessee, Knoxville, USA
Innovative Computing Laboratory
University of Victoria, Canada
School of Earth and Ocean Sciences
German Research School for Simulation Sciences, Germany
Laboratory for Parallel Programming
- Institut National de Recherche en Informatique et en Automatique, France
Barcelona Supercomputing Center
Computer Sciences - Computer Architecture for Parallel Paradigms
Tokyo Institute of Technology
Global Scientific Information and Computing Center
University of Tsukuba
Department of Computer Science