What began as an informal discussion two years between a physicist and a computer scientist has led to a new international collaboration aimed at creating computational tools to help scientists make more effective use of new computing technologies, including multicore processors. Among the scientific fields to be engaged are physics and astrophysics, climate modeling and geosciences.
The International Center for Computational Science (ICCS) is located at Lawrence Berkeley National Laboratory and the University of California at Berkeley, with partners at the University of Heidelberg in Germany and the National Astronomical Observatories of the Chinese Academy of Sciences in China. ICCS brings together expertise in hardware and programming as well as scientific disciplines to address the challenge of gathering, analyzing, storing and sharing the virtual explosion of scientific data resulting from experiments and simulations.
"Over the last decade we have seen a tremendous growth in computational science, and learned as a community that fastest progress will be made in large-scale collaborations," said Horst Simon, Associate Laboratory Director for Computing Sciences at Berkeley Lab and adjunct professor at UC Berkeley. "Therefore it is only natural to forge international collaborations to address the most challenging problems. ICCS is a step in this direction, combining projects on three continents to advance our scientific knowledge and the ability to use advanced architectures."
Many current scientific applications represent years of development and were developed for computer architectures based on single-core processors. But the next generation of supercomputers is based on multicore processors and will require new programming models and algorithms. Conversely, the new processor technologies are also powering remote scientific experiments, with systems using just one multi-core processor to balance the need for computing performance with low power consumption. A primary objective of ICCS is to develop scientific solutions that are easy to adopt, energy efficient and effective in implementation across a wide range of uses.
Even before the center was formally established, the partners had several key accomplishments. In December 2009, Hemant Shukla and John Shalf of LBNL and Rainer Spurzem of the University of Heidelberg organized a three-day international workshop as a part of an ongoing effort to develop infrastructure for enabling physics and astronomy data problems by utilizing emerging technologies. That meeting led to the international agreement resulting in ICCS.
"The center is a bridge between China, Germany and the USA. Through this bridge we will exchange knowledge, resources and personnel to work together on new technologies, as well as enhance our mutual cultural respect and understanding," said Spurzem, who has a joint appointment at the National Astronomical Observatories of the Chinese Academy of Science.
In April, a proposal by Shukla, Shalf and Horst Simon of Berkeley Lab and UC Berkeley received funding from the National Science Foundation (NSF) for their project known as ISAAC, or Infrastructure for Astrophysics Applications Computing. ISAAC is a three-year (2010-2013) funded project to focus on research and development of infrastructure to accelerate physics and astronomy applications using multicore architectures. The goalof the project is to successfully harness the power of the parallel architectures for compute-intensive scientific problems and open doors for new discovery and revolutionize the growth of science. More information can be found on the ISAAC web page.
"The entire HPC community is in transition towards a new paradigm for computation,"said Shalf, a Berkeley Lab computer scientist and center co-founder. "We think an application-driven approach to evaluating the alternative paths for future supercomputing software offers the best hope for creating a productive and durable environment that serves the needs of science."
The next steps for ICCS include hiring two researchers under the NSF project and to present a locally based summer session of the Virtual School of Computational Science and Engineering, a five-day course to learn how to use multicore devices for scientific computing, scale parallel code to tens of thousands, of CPU cores, handle large data volumes and more. Up to 35 students will attend the course in "Proven Algorithmic Techniques for Many-core Processors," presented in conjunction with the University of Illinois at Urbana-Champaign.
"Although high-performance computing for simulations, instrumentation and data analysis is a critical next step in scientific quests, the challenges of adopting HPC are steeped in unfamiliarity with rapidly changing technologies," said Hemant Shukla, a Berkeley Lab physicist and co-founder of the center. "Our goal is to enable broad range of scientific enquiry at various stages by developing real world solutions and engaging the wider scientific community, especially the younger generation."
For more information about the center, visit the ICCS website.
For more information about Computing Sciences at Berkeley Lab, please visit: