Parallel Model Order Reduction of Sparse Electromagnetic/Circuit Models

─ This paper describes a parallel Model Order Reduction (MOR) technique for Linear Time Invariant (LTI) electromagnetic/circuit systems with sparse structure. The multi-point Krylovsubspace projection method is adopted as framework for the model order reduction and a parallelization strategy is proposed. More specifically, a multi-point version of the wellknown PRIMA algorithm is proposed, which is parallelized with respect to the computation of the error between the original model and the reduced one. The number of moments to be matched for any expansion point is chosen adaptively as well. The numerical results show that the proposed parallelized MOR algorithm is able to preserve the accuracy of the reduced models while providing a significant compression and a satisfactory speedup with respect to the sequential one. Index Terms ─ Model order reduction, parallel computing, sparse electromagnetic/circuit systems.

[1]  Roland W. Freund,et al.  Recent Advances in Structure-Preserving Model Order Reduction , 2011 .

[2]  V. Simoncini,et al.  Block Krylov subspace methods for the computation of structural response to turbulent wind , 2011 .

[3]  William L. Goffe,et al.  Multi-core CPUs, Clusters, and Grid Computing: A Tutorial , 2005 .

[4]  Timothy A. Davis,et al.  Algorithm 832: UMFPACK V4.3---an unsymmetric-pattern multifrontal method , 2004, TOMS.

[5]  John N. Tsitsiklis,et al.  Parallel and distributed computation , 1989 .

[6]  Wha Wil Schilders,et al.  Introduction to model order reduction , 2008 .

[7]  Lawrence T. Pileggi,et al.  Asymptotic waveform evaluation for timing analysis , 1990, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[8]  M. Nakhla,et al.  Multipoint Full-Wave Model Order Reduction for Delayed PEEC Models With Large Delays , 2011, IEEE Transactions on Electromagnetic Compatibility.

[9]  Roger F. Harrington,et al.  Field computation by moment methods , 1968 .

[10]  Lawrence T. Pileggi,et al.  PRIMA: passive reduced-order interconnect macromodeling algorithm , 1997, ICCAD 1997.

[11]  A. Ruehli.  Equivalent Circuit Models for Three-Dimensional Multiconductor Systems , 1974 .

[12]  Jacob K. White,et al.  Automatic generation of accurate circuit models of 3-D interconnect , 1998 .

[13]  Boris Lohmann,et al.  How to choose a single expansion point in Krylov-based model reduction? , 2009 .

[14]  Michel S. Nakhla,et al.  Analysis of interconnect networks using complex frequency hopping (CFH) , 1995, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[15]  Gabriela Ciuprina,et al.  Effective extraction of accurate reduced order models for HF-ICs using multi-CPU architectures , 2012 .

[16]  Herng-Jer Lee,et al.  An adaptive-order rational Arnoldi method for model-order reductions of linear time-invariant systems , 2006 .

[17]  Albert E. Ruehli,et al.  The modified nodal approach to network analysis , 1975 .

[18]  J. Korvink,et al.  A Fully Adaptive Scheme for Model Order Reduction Based on Moment Matching , 2015, IEEE Transactions on Components, Packaging and Manufacturing Technology.