Research Publications

• A determinant theorem with applications to parallel algorithms, SIAM J. Numer. Anal., vol. 11 (1974), pp. 559-568.
• On the efficient computation of recurrence relations, ICASE, Hampton, Va.; Computer Science Dept., Carnegie-Mellon Univ., Pittsburgh, Pa., June 1974.
• Some aspects of the cyclic reduction algorithm for block tridiagonal linear systems, SIAM J. Numer. Anal., vol. 13 (1976), pp. 484-496.
• Accelerated iterative methods for the solution of tridiagonal systems on parallel computers, with D.K. Stevenson and J.F. Traub, J. ACM, vol. 23 (1976), pp. 636-654.
• Minimal parallelism for computations under time constraints, in High Speed Computer and Algorithm Organization, D.J. Kuck et al., eds., Academic Press, 1977, pp. 321-322.
• A survey of parallel algorithms in numerical linear algebra, SIAM Review, vol. 20 (1978), pp. 740-777.
• Minimal parallelism for associative computations under time constraints, Computing, vol. 22 (1979), pp. 101-118.
• On-line pipeline systems for recursive numeric computations, with M.J. Irwin, Proc. 7th Symp. on Computer Architecture, May 1980, pp. 292-299.
• Appendix to O. Wing and J.W. Huang, A computation model of parallel solution of linear equations, IEEE Trans. Computers, vol. C-29 (1980), pp. 632-638.
• More counterexamples to the LINPACK condition estimator, Computer Science Dept., Pennsylvania State Univ., July 1981.
• Optimal information for numerical differentiation, Computer Science Dept., Pennsylvania State Univ., Sept. 1981.
• Systolic networks for orthogonal equivalence transformations and their applications, with I.C.F. Ipsen, Proceedings of MIT Conf. on Advanced Research in VLSI, P. Penfield, ed., Artech House, Jan. 1982, pp. 113-122.
• Systolic networks for orthogonal decompositions, with I.C.F. Ipsen, SIAM J. Sci. Stat. Computing, vol. 4 (1983), pp. 261-269.
• Decomposition of recursive filters for linear systolic arrays, Real Time Signal Processing VI, K. Bromley, ed., Proc. SPIE, vol. 431, Aug. 1983, pp. 55-59.
• Multiprocessor Simulation Program SIMON, Shell Development Co., Houston, Nov. 1984.
• Mathematical hardware - design issues and responsibilities, in Algorithmically Specialized Parallel Computers, L. Snyder et al., eds., Academic Press, 1985, pp. 233-241.
• Partitioning big matrices for small systolic arrays, in VLSI and Modern Signal Processing, S.Y. Kung et al., eds., Prentice-Hall, 1985, pp. 185-199.
• Performance Measurements of the nCUBE/ten Multiprocessor, Shell Development Co., Houston, Dec. 1988.
• Ten Steps for Managing Parallel Computing Projects, IEEE Parallel & Distributed Technology, Spring 1994, pp. 6-8.
• Issues of Running Codes on Very Large Parallel Processing Systems, in Debugging and Performance Tuning for Parallel Computing Systems, M.L. Simmons et al., eds., IEEE Computer Society Press, 1996, pp. 233-243.

  • The suite of nCube system diagnostic programs described in this paper is available.

Shell Development Company, Internal Reports

• Simulation of Multiprocessor Systems for Geophysical Computing, with W.D. Moorhead, April 1984, 12 pp.
• Multiprocessor Simulation Program SIMON, July 1984, 62 pp. Edited version approved for public release.
• Multiprocessor Calculation of the Minimum-Time Algorithm, with W.D. Moorhead, Sept. 1984, 10 pp.
• Some Examples of Seismic Data Processing on a Panel-Oriented Parallel Computer, with J.S. Danburg and R.F. Starr, Nov. 1984, 12 pp.
• Hypercube Switching Element, March 1985, 9 pp.
• Two-Dimensional Fast Fourier Transforms for Hypercube Multiprocessors, July 1985, 33 pp.
• Experiments with SIMON: Bitonic Sorting on a Hypercube, March 1986, 48 pp.
• Software Technology, with W.S. Rhea, May 1987, 208 pp. A survey of the state-of-the-art in 1986, with projections 5-15 years hence.
• Basic Linear Algebra Subprograms for the nCUBE Node Processor, December 1987, 19 pp.
• A Short Course on Parallel Computing: The nCUBE Programming Environment, October 1988, 60 pp.
• A Tool for Analysis of Message-Passing Bugs on the nCUBE Multiprocessor System, November 1988, 14 pp.
• Performance Measurements of the nCUBE/ten Multiprocessor, December 1988, 28 pp. Edited version approved for public release.
• Production Use of Parallel Computing Within Shell Oil, with 19 co-authors, 1989 Geophysical Conference, Jan. 1989, 26 pp.
• GASSIM on the nCUBE: Application of a Parallel Computer to Reservoir Simulation, 1989 Production Computer Systems Conference, March 1989, 4 pp.
• Parallel Computing Debugging Tools: An Instrumented Operating System for the nCUBE Multiprocessor, Advanced Technology Workshop, Bellaire Research Center, November 1989, 20 pp.
• Software Portability Among Shell's Parallel Computer Systems, KSEPL report, Oct. 1991, 47 pp.
• Intel Parallel Computer Systems, Nov. 1992, 50 pp.
• Fundamental Advice for Management of an Alternative Architecture Project, Sept. 1993, 4 pp.

Shell Development Company, Significant Software Developments

• Simulation studies that led to selection of nCUBE systems for seismic data processing.
• Continual beta-test of nCUBE products, 1985-1993.
• Ported a single-phase reservoir simulator to the nCUBE-1, advised and participated in later efforts by others at BRC and KSEPL.
• Software quality assurance tools for the nCUBE-1 and nCUBE-2; instrumented OS with message-passing bug analyzer, OS data structure scanner.
• Consultant studies, performance analysis and program modifications for a wide variety of scientific applications.

Invited Lectures at Conferences

• A Simulator for Multiprocessor Networks, ARO Workshop, New Computing Environments: Parallel, Vector and Systolic, Stanford Univ., Nov. 7-9, 1984.
• Early Experience with the nCUBE at Shell, Second Conference on Hypercube Multiprocessors, Knoxville, Sept. 29-30, 1986.

Panel Discussions at Conferences

• Hypercube Simulators, First Conference on Hypercube Multiprocessors, Knoxville, Aug. 26-7, 1985.
• Computer Technology and Seismic Modeling Workshop, Soc. Exploration Geophysicists, Houston, Nov. 6, 1986.
• Are advances in supercomputing occurring rapidly enough?, Second Int. Conf. of Supercomputing, Santa Clara, May 5-8, 1987.
• Massively Parallel vs. Parallel Vector Supercomputers: a User's Perspective, Supercomputing '93, Portland, Ore., Nov. 15-19, 1993. (abstract in conference proceedings, pp. 918-20)

Invited Lectures at universities, etc.

• Sandia Laboratories, Albuquerque, 1979, 1982
• Los Alamos Scientific Laboratory, 1979
• IBM T.J. Watson Research Center, 1981
• University of New Mexico, 1982
• University of Minnesota, 1982
• Rice University, 1983 (IEEE ASSP, Houston chapter)
• Yale University, 1985
• Pennsylvania State University, 1987
• University of Houston, 1988
• University of Texas, Austin, 1988
• Pittsburgh Supercomputing Center, 1992
• Minnesota Project Innovation, 1994

Gordon Bell Prize selection committee, 1993-4

• Gordon Bell Prize Lectures 1993, Proc. Supercomputing '93, pp. 514-5.
• Summary articles in IEEE Computer, Jan. 1994, pp. 69-75, and Jan. 1995, pp. 68-74.