RELATED WEBSITE: Emeagwali's Website

Philip Emeagwali has given real meaning to the NSFNET expression"breaking down the barriers of space and time." Gordon Bell Prize winner Emeagwali, a Ph.D. candidate in the University of Michigan's Department of Civil Engineering and Program in Scientific Computing, utilized the vast resources of the NSFNET to win first place in the 1989 Gordon Bell Prize Competition. PHILIP EMEAGWALI "Fastest supercomputer on earth" The $1000 prize for this competition, known as the "supercomputer olympics," recognizes outstanding achievement in the use of supercomputers to solve significant scientific and engineering problems. The Institute for Electrical and Electronics Engineers (IEEE) competition is considered the annual high point of supercomputer research. Each year, the winning supercomputer is declared the "fastest supercomputer on earth." This is the first time the prize has been given to a sole investigator and the first time it has been given to an individual from a university. All previous IEEE Gordon Bell Prize Competition winning entries were collaborative efforts, involving researchers from industry, academia, and research laboratories. Previous participants have included multidisciplinary teams from Mobile Oil, Cray Research, IBM, the California Institute of Technology, MIT, the NASA Ames Research Center, the National Center for Atmospheric Research, AT&T Bell Labs, and other institutions. Underground petroleum recovery Emeagwali's prize-winning research focuses on underground petroleum recovery through high-speed supercomputer simulations. His primary research interests are parallel computation and large- scale problems in computational fluid dynamics. His research was undertaken on Connection Machine supercomputers at the Los Alamos National Laboratory at Los Alamos, New Mexico, the Argonne National Laboratory at Argonne, Illinois, the National Center for Supercomputer Applications at the University of Illinois at Urbana- Champaign, and the Thinking Machines Corporation in Cambridge, Massachusetts. The NSFNET enabled Emeagwali to access the supercomputers from his local workstation. "I have more than a dozen accounts, on a dozen different computers, in a dozen different cities," Emeagwali explained. "I have never been to these cities, but through NSFNET, I was able to access every one of these machines." NSFNET essential "NSFNET was absolutely necessary for me to perform this work. As a matter of fact, without NSFNET, I never would have been able to conduct my research," Emeagwali said when explaining the importance of the network. "The network greatly improved my productivity by facilitating my collaboration with my colleagues at various locations," he added. Background of study In 1988, the cost of imported oil accounted for 29 percent of the United States trade deficit. This fact, coupled with the drive toward less dependence on imported oil, makes it important to maximize the amount of oil recovered from petroleum production wells. Currently, engineers can only recover about 30 percent of the oil in a petroleum reservoir. By using petroleum reservoir simulation models to manage a group of oil wells economically, engineers can recover more oil. Because supercomputers are often used to solve the equations used in petroleum reservoir simulations and increase the total amount of recoverable oil, it is not surprising that 10 percent of supercomputers in existence have been purchased by the petroleum industry. Grand challenge Given the huge economic benefits to be derived and the fact that more powerful supercomputers are needed for accurate reservoir simulation, such simulations have been designated by the U.S. government as one of the 20 national Grand Challenges in science and engineering. Connection Machine Supercomputer Emeagwali conducted his prize-winning research on the Connection Machine supercomputer. The Connection Machine, one of the fastest supercomputers ever built, consists of a collection of more than 65,000 separate processors cooperating simultaneously to solve single, complex problems. The Connection Machine is ideal for applications that require the simultaneous performance of thousands and even millions of simple arithmetical operations. For such computation-intensive applications, the processing power of the Connection Machine actually increases as the amount of data increases. NSFNET Reliability Even though he did not have a Connection Machine in his own backyard, NSFNET brought the resources of this powerful supercomputer to his desktop workstation. Emeagwali commented on the reliability and convenience of the network: "The network was extremely reliable. I was able to conduct my research at any time during the day or night." Challenges The Connection Machine simulation of petroleum reservoirs poses several mathematical and programming challenges: formulate a set of governing equations that adequately describes the flow behavior of oil in a reservoir and yields algorithms suitable for the Connection Machine architecture reduce the inter-processor communication time in the Connection Machine by creating several million virtual (or "artificial") processors decompose and evenly distribute the workload among the several million virtual processors created. Currently, few algorithms are suitable for the architecture of supercomputers like the Connection Machine. Only a few complex, real-life problems can be solved on such machines. Many other problems would be potentially solvable if appropriate algorithms could be developed. Effective supercomputer use To use the newer supercomputers effectively, researchers must rethink, rewrite and reformulate many computation-intensive applications. New supercomputers will stimulate the development of new problem-solving approaches, new governing equations (or descriptions) for important problems, and new numerical algorithms. Developing the algorithm The radically different architecture of the Connection Machine motivated Emeagwali to design and implement a different and more suitable algorithm for petroleum reservoir simulation.  RELATED ARTICLES   Computer Wizard   Emeagwali Puts Math to Work in Real World   Top computing prize goes to a University of Michigan Ph.D. candidate   The Ways of Counting   Grad Student Wins Top Supercomputer Award   One of the World's Fastest Humans   Philip Emeagwali wins 1989 Gordon Bell Supercomputing Prize   U-M Researcher wins Gordon Bell Prize   Brain to Brain (Detroit Free Press)   New Stars Emerging in Adult Tennis Tourney   Award Winner Cites NSFNET Reliability   The answer is in the wind   Please visit http://emeagwali.com for the most recent list. He obtained his algorithm by modifying a set of governing equations developed in 1938 by the Russian mathematician B.K. Risenkampf. Although the equations were abandoned for various historical and computational reasons, Emeagwali argues that they are suitable for the newer supercomputers such as the Connection Machine. More importantly, his approach is applicable to a wide range of important scientific and engineering problems, including the problems of calculating the movement of buried nuclear wastes. Another challenge associated with the use of the Connection Machine is the time spent in inter-processor communication, which had made it extremely difficult to obtain very high performance. Emeagwali was able to reduce inter-processor communication time drastically by creating and using more than eight million virtual processors instead of the original 65,000 processors. Results-3.1 billion calculations/sec Using this new approach in combination with the Connection Machine, Emeagwali's model ran at the exceptionally high speed of 3.1 billion calculations per second-three times faster than the 1988 Gordon Bell Prize winner and 24 times faster than the 1987 winner. The speed of Emeagwali's model even exceeds the theoretical peak calculation speed of much more expensive conventional supercomputers, including the widely used $30 million Cray Y-MP. From several hours to a few seconds Running at such speeds, petroleum reservoir simulation problems that formerly took several hours to solve on conventional supercomputers can now be solved in only a few seconds. Reported in the Link Letter of the Merit Network, Inc. (and National Science Foundation) in May/June 1990. Click on this icon to Click on emeagwali.com for more information.