What are IU faculty and leaders saying about Big Red II?
Leaders and Researchers in Medicine and Biology
This new computer is a dream come true for any computational scientist and strengthens IU's long standing tradition of being one of, if not the best University in the world for computational scientists. My research group has been studying how we can promote reactions that will turn water, carbon dioxide and solar energy into chemical fuel by carrying out large-scale computer simulations. These computer simulations of fundamental processes of artificial photosynthesis pose daunting demands on the computational power. With the new machine, I expect that we will be able to attempt to design more realistic simulations that will help us understand what we need to do to make the dream of artificial photosynthesis come true in the future.
Having been involved in the evolution of IU's advanced computing environment since the beginning of INGEN in 2000, I have seen how advanced computing has become more and more critical to medical research and innovation, and watched as the IU computational resources have been deployed in ways that are more and more valuable to IU medical research. Big Red II will be a critical and strategic aid to accelerating new medical breakthroughs and enabling research that will improve human health.
Data sets of unprecedented scope can facilitate new discoveries regarding the brain, genome, disease and therapies but computational power has become a major bottleneck to scientific progress. To analyze the entire human genome in relation to longitudinal changes on brain MRI and PET scans in over 800 individuals we need an order of magnitude more computing power than available today. The new [supercomputer] is an exciting development that will undoubtedly enable new discoveries my many investigators at IU and beyond.
Leaders and researchers in the natural and physical sciences
The Big Red II computer that Indiana University will purchase is the most exciting development in campus computing in well over a decade. The GPU based nodes are very forward looking and will provide a big boost in terms of energy consumption and cost effectiveness for codes that can make use of them. My research will definitely benefit from this new approach.
In 2009-2010, I spent my sabbatical year at the National Center for Supercomputing Applications (NCSA) to prepare for the new Blue Waters computer that will be the National Science Foundation's premier computing platform. The development of GPU capability for our code is one of the most useful results of my visit. The similarity of Big Red II to Blue Waters will help IU researchers to prepare for working on Blue Waters.
As head of the Ph.D. minor in Scientific Computing I have been requesting that we offer a course that will teach our students how to write code for GPUs. Big Red II will be an excellent platform on which students can develop their codes and carry out cutting edge research.
Big Red II will be a great help to my students and me in advancing our Research.
I am excited to learn of Indiana University's plan for a new Cray supercomputer. This machine should significantly aid my group's research performing large scale molecular dynamics simulations of white dwarf and neutron star material. Our simulations, performed by a full range of high school, undergraduate, graduate, and post graduate students, help determine the properties of extremely dense matter. For example, our finding that neutron star crust is the strongest material known may facilitate an important test of Albert Einstein's General Theory of Relativity. In 2012 we will use about five million core hours on Kraken, the Cray XT5 supercomputer at Oak Ridge Tennessee. Because of its similar architecture, the new Indiana machine should be particularly useful to us. Not only will the new supercomputer be an important tool for advancing astrophysics, it will also provide an outstanding laboratory for training the next generation of computational physicists.
Patterns in turbulence emerge only upon taking long time averages, and resolving small length scales. Our mathematical investigations both suggest and are guided by direct numerical simulations. A large super computer would enable us to explore much finer scales in both 2-D and 3-D flows. The fundamental study of turbulence impacts a wide range of applications, including aerospace engineering, weather forecasting, and blood circulation, just to name a few.
Given greatly enhanced supercomputing resources, the research group led by Distinguished Professor Peter Ortoleva will be able to make predictions of the behavior of viruses and other microbes from the basic laws of physics. The biotechnical consequences of this predictive capability will be a computer-aided procedure for the efficient design of vaccines against viral threats to global health and security.
The new supercomputer will be a great benefit for research in astronomy, both for our expanding computational theory program in stellar dynamics and for handling the data from our new, large format camera on the WIYN telescope. The new supercomputer will make this work fly!
Members of my lab have over the years used almost all of IU's computational resources in our work: Quarry, BigRed, Mason, and storage facilities. Many of our projects in genomics and population biology are computationally demanding, and a replacement of the BigRed cluster will certainly be much appreciated, and enthusiastically used. Generally, we study the mechanisms of evolution at the gene, genomic, cellular, and phenotypic levels, with special attention being given to the roles of mutation, random genetic drift, and recombination. These processes underlie everything from the spread of genetic diseases in humans, to conservation biology of endangered species.
Leaders and Researchers in Informatics and Computing – including the IU Pervasive Technology Institute
This new system will enable the Sustained Environments – Actionable Data project to advance research in large-scale scientific data sharing and continue Nobel Laureate Lin Ostrom’s legacy in sustainability science research. Big Red II will also enable the Hathi Trust Research Center to tackle the BigData challenge of new methodologies for analyzing and enhance the 10M+ digitized books that make up the HathiTrust corpus.
The scientific study of the structure and dynamics of science and technology is critical for economic and societal progress. Researchers at IU and world wide are developing and validating analytic and predictive models of the science and innovation system to support decision making in research, industry, education, and science policy making. IU is the home of key databases and tool development efforts (e.g., the Scholarly Database (sdb.cns.iu.edu), the DIRECT Network (direct2experts.org), Scholarometer.indiana.edu), and plug-and-play macroscopes such as the Science of Science (Sci2) Tool (sci2.cns.iu.edu) and the Big Red II will be instrumental for mining, simulating, and visualizing massive amounts of scholarly and social media data streams using a mixed methods, multi-level approach.
Big Red II puts IU at the forefront of academic high performance computing research spanning the largest scientific simulations, data mining with new emerging algorithms and core computer science research including new languages and runtimes that will enable exascale systems.
One of the major challenges for computational molecular biologists is computer power. Molecular dynamics simulation that requires high computational capacity can help visualize and understand the behavior and function of macromolecules such as proteins, protein-protein complexes and protein-RNA complexes and therefore expand our understanding of fundamental biological processes such as aging and diseases at molecular level. Modern graphics processing units (GPUs) can provide tremendous acceleration for numerous intensive scientific applications including molecular modeling and dynamics. The new supercomputer will surely enable investigators at IU to initiate more computational expensive projects leading to cutting-edge scientific discoveries that may not happen if without the resources.
This new acquisition continues IU's long tradition of leadership in supercomputing. By being able to solve much larger and much more complex problems than they ever could before, researchers at IU will be able to open up entirely new scientific vistas. CREST (the Center for Research in Extreme Scale Technologies) is excited to be able to contribute to such a vital scientific resource
The School of Informatics and Computing’s thriving research programs, supported by a record level of new grant funding last year, is greatly aided by IU’s excellent computing resources. This new supercomputer will enhance the significance and scope of that research and create new research and education opportunities for our School. In addition, SoIC researchers, in good part through their involvement in the Pervasive Technology Institute, will play a key role in developing the software and tools that will enable other IU researchers to get the greatest possible value from this historic investment in high performance computing at IU.
Big Red II will enable the Center for Research in Extreme Scale Technologies to dramatically advance its research in big-data oriented dynamic graph processing. These highly irregular time-varying data structures challenge conventional scale systems and the exceptional Gemini network technology combined with the very high scale of computational nodes will permit first rate symbolic data processing. As this system is of the same family of systems being deployed both at NCSA and ORNL, it provides a direct and seamless migration path to systems among the most powerful in the world.
My research group studies how to efficiently utilize heterogeneous parallel computing resources. The new HPC system presents an ideal test-bed for this work. It will not only include a large number of nodes, but many CPU cores per node, imbalanced floating point and integer units, and mixed CPU/GPU nodes. This system will be both a challenge for and a significant boost to our research on parallel scheduling algorithms.
Leaders in Office of the Vice President for Research and University Information Technology Services, including the Pervasive Technology Institute
I have been part of the IU community for more than 30 years. This is one of the most exciting steps in IU research at that time. For IU information technologists and scientists to participate in finding clear evidence of the Higgs Boson and to enter into this strategic partnership in Cray in one year is truly phenomenal. IU is forging new paths in enabling important new research discoveries at IU with a 1 PetaFLOPS supercomputer for use by our university community. I expect that we will see even more new breakthroughs than we have seen with the original Big Red.
The new research opportunities and scales of data analysis made possible by Big Red II will be energizing for IU researchers and UITS research support experts alike. For many years, our science gateways,visualization, and data analysis teams have focused on scalable methods and software tools for use with the national cyberinfrastructure.
It will be very rewarding to apply more of our world-class expertise back home again in Indiana, and the transition to Big Red II should prove to be rapid and highly impactful.
The partnership with Cray represents IU’s continued excellence supporting research in higher education. I am proud to be a part of an organization where high performance computing is an important part of the IT strategic plan. Big Red II will be an immense resource to the research community and I anxiously await the discoveries that will follow.
I am excited for IU researchers to have local access to a Cray supercomputer. Big Red II will provide a great balance between running large scale parallel workloads that take full advantage of the Gemini interconnect, while maintaining compatibility with standard Linux applications through Cluster Compatibility Mode. I am looking forward to working with researchers to optimize and tune their scientific applications to take full advantage of the new system.