Abstract
We describe an approach to the parallel and distributed solution of large-scale, block structured semidefinite programs using the spectral bundle method. Various elements of this approach (such as data distribution, an implicitly restarted Lanczos method tailored to handle block diagonal structure, a mixed polyhedral-semidefinite subdifferential model, and other aspects related to parallelism) are combined in an implementation called LAMBDA, which delivers faster solution times than previously possible, and acceptable parallel scalability on sufficiently large problems.
Similar content being viewed by others
References
Benson S.J., Ye Y., Zhang X. (2000) Solving large-scale sparse semidefinite programs for combinatorial optimization. SIAM J. Optim. 10(2): 443–461
Burer S. (2003) Semidefinite programming in the space of partial positive semidefinite matrices. SIAM J. Optim. 14(1): 139–172
Burer, S., Monteiro, R.D.C.: A nonlinear programming algorithm for solving semidefinite programming via low-rank factorization. Math. Program. (Ser. B) 95, 329–357 (2003)
Burer, S., Monteiro, R.D.C.: Local minima and convergence in low-rank semidefinite programming. Math. Program. (2006)(to appear)
Cullum J., Donath W., Wolfe P (1975). The minimization of certain nondifferentiable sums of eigenvalues of symmetric matrices. Math. Program. Study 3: 35–65
Fletcher R. (1985) Semidefinite matrix constraints in optimization. SIAM J. Control Optim. 23: 493–523
Fujisawa, K., Kojima, M., Nakata, K., Yamashita, M.: SDPA User’s Manual – Version 6.00. Department of Mathematical and Computing Sciences, Tokyo Institute of Technology (2002)
Fukuda M., Kojima M., Murota K., Nakata K. (2000) Exploiting sparsity in semidefinite programming via matrix completion I: general framework. SIAM J. Optim. 11(3): 647–674
Fukuda, M., Braams, B.J., Nakata, M., Overton, M.L., Percus, J.K., Yamashita, M., Zhao, Z.: Large-scale semidefinite programs in electronic structure calculations. Technical report B-413, Department of Mathematical and Computing Sciences, Tokyo Institute of Technology (2005)
Helmberg, C.: SBmethod: a C++ implementation of the spectral bundle method. Technical report ZR-00-35, Konrad-Zuse-Zentrum für Informationstechnik, Berlin (2000)
Helmberg, C.: Semidefinite programming for combinatorial optimization. Technical report ZR-00-34, TU Berlin, Konrad-Zuse-Zentrum, Berlin (2000)
Helmberg C. (2003) Numerical evaluation of SBmethod. Math. Program. 95(2): 381–406
Helmberg C., Kiwiel K.C. (2002) A spectral bundle method with bounds. Math. Program. 93(2): 173–194
Helmberg C., Rendl F. (1999) A spectral bundle method for semidefinite programming. SIAM J. Optim. 10(3): 673–696
Hiriart–Urruty J–B., Lemaréchal C. (1993) Convex Analysis and Minimization Algorithms, vol. I & II. Springer, Berlin Heidelberg New York
Hiriart–Urruty J–B., Ye D. (1995) Sensitivity analysis of all eigenvalues of a symmetric matrix. Numer. Math. 70: 45–72
Kiwiel K.C. (1983) An aggregate subgradient method for nonsmooth convex minimization. Math. Program. 27: 320–341
Kiwiel K.C. (1990) Proximity control in bundle methods for convex nondifferentiable minimization. Math. Program. 46:105–122
Kočvara, M., Stingl, M.: On the solution of large-scale SDP problems by the modified barrier method using iterative solvers. Technical report 304, Institute of Applied Mathematics, University of Erlangen (2005)
Lehoucq R.B., Sorensen D.C., Yang C. (1998) ARPACK Users’ Guide. SIAM, Philadelphia
Mazziotti D.A. (2004) First-order semidefinite programming for the direct determination of two-electron reduced density matrices with application to many-electron atoms and molecules. J. Chem. Phys. 121: 10957–10966
Mazziotti D.A. (2004) Realization of quantum chemistry without wavefunctions through first-order semidefinite programming. Phys. Rev. Lett. 93: 213001
Mittelmann H.D. (2003) An independent benchmarking of SDP and SOCP solvers. Math. Program. 95: 407–430
Nakata K., Fujisawa K., Fukuda M., Kojima M., Murota K. (2003) Exploiting sparsity in semidefinite programming via matrix completion II: implementation and numerical results. Math. Program. 95(2): 303–327
Nayakkankuppam, M.V.: Optimization over symmetric cones. Ph.D. Thesis, New York University (1999)
Nayakkankuppam, M.V., Tymofyeyev, Y.: A parallel implementation of the spectral bundle method for semidefinite programming. In: Proceedings of the 8th SIAM Conference on Applied Linear Algebra. SIAM, Williamsburg (2003)
Overton M.L. (1988) On minimizing the maximum eigenvalue of a symmetric matrix. SIAM J. Matrix Anal. Appl. 9(2): 256–268
Parlett B.M. (1998) The Symmetric Eigenvalue Problem. SIAM, Philadelphia
Rockafellar R.T. (1970) Convex Analysis. Princeton University Press, Princeton
Sorensen D.C. (1992) Implicit application of polynomial filters in a k-step Arnoldi method. SIAM J. Sci. Comput. 13(1): 357–385
Toh K.C. (2004) Solving large scale semidefinite programs via an iterative solver on the augmented systems. SIAM J. Optim. 14(3): 670–698
Toh K.C., Kojima M. (2002) Solving some large scale semidefinite programs via the conjugate residual method. SIAM J. Optim. 12(3): 669–691
Yamashita, M., Fujisawa, K., Kojima, M.: SDPARA: Semi Definite Programming Algorithm: paRAllel version. Parallel Comput. 29, 1053–1067 (2003). http://grid.r.dendai.ac.jp/sdpa
Zhao Z., Braams B.J., Fukuda M., Overton M.L., Percus J.K. (2004) The reduced density matrix method for electronic structure calculations and the role of three-index representability conditions. J. Chem. Phys 120(5): 2095–2104
Author information
Authors and Affiliations
Corresponding author
Additional information
Dedicated to the memory of Jos Sturm.
This work was supported in part by NSF grants DMS-0215373 and DMS-0238008.
Rights and permissions
About this article
Cite this article
Nayakkankuppam, M.V. Solving large-scale semidefinite programs in parallel. Math. Program. 109, 477–504 (2007). https://doi.org/10.1007/s10107-006-0032-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10107-006-0032-1
Keywords
- Semidefinite programming
- Eigenvalue optimization
- Subgradient bundle methods
- Lanczos method
- Parallel computing