Please start by filling out the survey

About the Course

• CS 395T
  Parallel Techniques for Numerical Algorithms
• Unique Number: 47470
• Time: MW 3:30-4:45
• Room: GAR 201

About the Professor

• Robert A. van de Geijn
• Associate Professor of Computer Sciences
• Office: Tay 4.115C
• e-mail: rvdg@cs.utexas.edu
• phone: 471-9720
• Office Hours: to be determined

Materials

• Required:
  MPI: THE COMPLETE REFERENCE by Snir, Otto, Huss-Lederman, Walker, and Dongarra, MIT Press, 1995.
  DO NOT PRINT THESE BOOKS OUT.
• Reference:
  Using MPI by Gropp, Lusk, and Skjellum, MIT Press, 1994
• Other:
  Other materials will be place on the webpage http://www.cs.utexas.edu/users/rvdg/395T.96

Syllabus

• Why Parallel Computing?
• Why Message Passing?
• Collective Communication
• High Performance Parallel Dense Linear Algebra Algorithms
• Sparse Methods
• Selected Topics

Grading

Reference Information

The Argonne National Laboratory MPI site.

Information on various vendors

• Intel Scalable Systems Division Intel Paragon
• IBM MPI on SP-2.
• Cray T3D and T3E
• Silicon Graphics Incorporated Power Challenge
• Convex SPP

Various documents used in the class

• A Street Guide to Collective Communication on Parallel Computers
  • Chapter 1: Handout, Jan. 22
  • Chapter 1: Handout, Jan. 24
  • Chapter 1-2: Handout, Jan. 29
• BLAS materials
  • BLAS quick reference guide. (Hardcopies handed out in class)
  • Level 2 BLAS paper
  • Level 3 BLAS paper
• Using the iPSC/860 See James Overfelt's page
• Communication Materials
  • NEW Prasenjit Mitra, David Payne, Lance Shuler, Robert van de Geijn, and Jerrell Watts, "Fast Collective Communication Libraries, Please," Proceedings of the Intel Supercomputing Users' Group Meeting 1995.

  • Hypercube Global Combine Paper
  • MESH Global Combine Paper
  • Barnett, et. al. ``Broadcasting on Meshes with Wormhole Routing,'' Journal of Parallel and Distributed Computing
  • Jerrell Watts and Robert van de Geijn, Mesh Pipelined Broadcast Paper
  • M. Barnett, S. Gupta, D. Payne, L. Shuler, R. van de Geijn and J. Watts, ``Interprocessor Collective Communication Library (InterCom),'' Scalable High Performance Computing Conference 1994.
• Matrix-vector Multiplication Materials
  • Output from matrix-vector example
  • SHPCC94 MV multiply paper
  • Manuscript on MV multiplication
• Matrix-Matrix Multiplication Materials
  • NEW Almadena Chtchelkanova, John Gunnels, Greg Morrow, James Overfelt, Robert A. van de Geijn, "Parallel Implementation of BLAS: General Techniques for Level 3 BLAS," submitted to Concurrency: Practice and Experience

  • NEW Robert van de Geijn and Jerrell Watts "SUMMA: Scalable Universal Matrix Multiplication Algorithm," Concurrency: Practice and Experience , to appear.

  • NEW Brian Grayson and Robert van de Geijn "A High Performance Parallel Strassen Implementation," Parallel Processing Letters, to appear.

  • Choi et al.: PUMMA paper
  • Huss-Lederman, S., E. Jacobson, A. Tsao, & G. Zhang, "Matrix Multiplication on the Intel Touchstone DELTA," Technical Report SRC-TR-93-101 (revised), Supercomputing Research Center, (Feb. 1994, 21 pages). This is an expanded version of PRISM Working Note #7 and supersedes PRISM Working Note #11. Appeared in Concurrency: Practice and Experience, Vol. 6 (7), Oct. 1994, pp. 571-594.
• Dense Factorization Materials
  • Dongarra, van de Geijn, and Walker, "Scalability Issues Affecting the Design of a Dense Linear Algebra Library" Journal of Parallel and Distributed Computing, Vol. 22, No. 3, Sept. 1994.
  • LAPACK LU factorization codes
    • dgetrf.f (level 3)
    • dgetf2.f (level 2)
  • Parallel LU factorization codes
    • pzlubr.f (level 3)
    • pzlur.f (level 2)
• Eigenvalue Materials
  • Jack Dongarra and Robert van de Geijn ``Reduction to Condensed Form on Distributed Memory Architectures,'' LAPACK Working Note 30, in Parallel Computing , 18, pp. 973--982, 1992.
  • J. Choi, J. Dongarra, and D. Walker, ``The Design of a Parallel Dense Linear Algebra Software Library: Reduction to Hessenberg, Tridiagonal, and Bidiagonal Form,'' UT, CS-95-275, February 1995.