An introduction to high performance computing: technology, tools and applications

• Autorzy: Kowalik J. S.
• Języki publikacji: EN

Abstrakty

EN

Traditionally high performance was applied to very computationally demanding problems in science and engineering. They were known as the Grand Challenge Problems that required supercomputers equipped with very large computer memories and capable of high rates of computation measured by Megaflops (million floating point operations per second). In the last two decades the world of business and industry has recognized the enormous potential of large distributed computer systems for their business enterprise applications. Today most of the existing high performance computers are employed in the enterprise environments. A typical system has three layers of servers: user interface, applications and database. These client/server architectures are the work horse of large enterprise information processing. This introductory paper describes both areas of high performance computing applications that differ in the nature of workload, performance objectives, design methodology and scientific focus.

Słowa kluczowe

EN

supercomputers; client-server systems; enterprise data processing systems; performance; parallel computing; cluster computing

• Wydawca: Politechnika Gdańska
• Czasopismo: TASK Quarterly : scientific bulletin of Academic Computer Centre in Gdansk
• Rocznik: 2000
• Tom: Vol. 4, No 2
• Strony: 171--178
• Opis fizyczny: Bibliogr. 7 poz., rys.

Twórcy

autor

Kowalik J. S.

• The Boeing Company, PO Box 3707, MS 7L-44, Seattle, Washington 98124-2207, USA
• University of Washington, Department of Computer Science and Engineering, FR-35, Seattle, Washington, Seattle 98195, USA

Bibliografia

• [1] Gropp W., Lusk F. and Skjelllum A., Using MPI: Portable Parallel Programming with the Message Passing Interface, The MIT Press, Sc:,’ntific and Engineering Computation Series, 1994
• [2] How to Build Beowulf The MIT Press, Scientific and Engineering Computation Series, 1999
• [3] Dongarra J., Duff I. and Van der Vorst H. A., Numerical Linear Algebra for High-Performance Computing, SIAM Press, 1998
• [4] Gustafson J., Reevaluation of Amdahl's Law, Comm. ACM, 31,531, 1988
• [5] Kuck D. J., High Performance Computing, Oxford Umversny Press, 1996
• [6] Aries J., Banerjee S., Brittan M., Dillon E., Kowalik J. and Lixvar J., Capacity and Performance Analysis of Distributed Enterprise Systems, submitted to Comm, of ACM
• [7] Nezlek G. S., Jain H. K. and Nazareth D. L., An Integrated Approach to Enterprise Computing Architecture, Comm. ACM, 42,82,1999