Parallelizing the QR Algorithm for the Unsymmetric Algebraic Eigenvalue Problem: Myths and Reality

Over the last few years, it has been suggested that the popular QR algorithm for the unsymmetric Schur decomposition does not parallelize. In this paper, we present both positive and negative results on this subject. In theory, asymptotically perfect speedup can be obtained. In practice, reasonable speedup can be obtained on an MIMD distributed memory computer for a relatively small number of processors. However, we also show theoretically that it is impossible for the standard QR algorithm to be scalable. Performance of a parallel implementation of the LAPACK DLAHQR routine on the Intel ${\text{Paragon}}^{{\text{TM}}} $ system is reported.

[1]  Jack J. Dongarra,et al.  A Parallel Algorithm for the Reduction of a Nonsymmetric Matrix to Block Upper-Hessenberg Form , 1995, Parallel Comput..

[2]  Robert A. van de Geijn Implementing the qr-algorithm on an array of processors , 1987 .

[3]  R. Byers Numerical Stability and Instability in Matrix Sign Function Based Algorithms , 1986 .

[4]  Daniel Boley,et al.  Solving the generalized eigenvalue problem on a synchronous linear processor array , 1986, Parallel Comput..

[5]  Jack J. Dongarra,et al.  A Parallel Algorithm for the Nonsymmetric Eigenvalue Problem , 1991, SIAM J. Sci. Comput..

[6]  James Demmel,et al.  On a Block Implementation of Hessenberg Multishift QR Iteration , 1989, Int. J. High Speed Comput..

[7]  Greg Henry Increasing Data Reuse in the Unsymmetric QR Algorithm , 1992 .

[8]  Ilse C. F. Ipsen,et al.  Complexity of dense linear system solution on a multiprocessor ring. Research report , 1986 .

[9]  J. H. Wilkinson The algebraic eigenvalue problem , 1966 .

[10]  James Demmel,et al.  Design of a Parallel Nonsymmetric Eigenroutine Toolbox, Part I , 1993, PPSC.

[11]  Daniel Boley,et al.  A parallel QR algorithm for the nonsymmetric eigenvalue problem , 1989 .

[12]  Bruce Hendrickson,et al.  The Torus-Wrap Mapping for Dense Matrix Calculations on Massively Parallel Computers , 1994, SIAM J. Sci. Comput..

[13]  Gregory Mark Henry Improving data re-use in eigenvalue-related computations , 1994 .

[14]  Robert A. van de Geijn,et al.  Reduction to condensed form for the eigenvalue problem on distributed memory architectures , 1992, Parallel Comput..

[15]  Al Geist,et al.  Finding eigenvalues and eigenvectors of unsymmetric matrices using a distributed-memory multiprocessor , 1990, Parallel Comput..

[16]  J DongarraJack,et al.  Scalability issues affecting the design of a dense linear algebra library , 1994 .

[17]  L. Auslander,et al.  On parallelizable eigensolvers , 1992 .

[18]  Robert A. van de Geijn,et al.  Storage Schemes for Parallel Eigenvalue Algorithms , 1988 .

[19]  Ilse C. F. Ipsen,et al.  The Impact of Parallel Architectures on The Solution of Eigenvalue Problems , 1986 .

[20]  David S. Watkins,et al.  Shifting Strategies for the Parallel QR Algorithm , 1994, SIAM J. Sci. Comput..

[21]  G. W. Stewart,et al.  A parallel implementation of the QR-algorithm , 1987, Parallel Comput..

[22]  G. A. Geist,et al.  Finding eigenvalues and eigenvectors of unsymmetric matrices using a hypercube multiprocessor , 1989, C3P.

[23]  J. Demmel Open problems in numerical linear algebra , 1992 .

[24]  Patricia J. Eberlein,et al.  On the Schur Decomposition of a Matrix for Parallel Computation , 1985, IEEE Transactions on Computers.

[25]  J. G. F. Francis,et al.  The QR Transformation A Unitary Analogue to the LR Transformation - Part 1 , 1961, Comput. J..

[26]  Steven Huss-Lederman,et al.  A Parallelizable Eigensolver for Real Diagonalizable Matrices with Real Eigenvalues , 1997, SIAM J. Sci. Comput..

[27]  Christian H. Bischof,et al.  A Parallel Algorithm for the Sylvester Observer Equation , 1996, SIAM J. Sci. Comput..

[28]  Greg Henry Improving the Unsymmetric Parallel QR Algorithm on Vector Machines , 1993, PPSC.

[29]  James Demmel,et al.  The Spectral Decomposition of Nonsymmetric Matrices on Distributed Memory Parallel Computers , 1997, SIAM J. Sci. Comput..