Title :
Parallelization of the Reduced-Coupling Technique for a Method-of-Moments-Based Field Solver Used for Product-Level Wide Data-Bus Analysis
Author :
Hesford, A.J. ; Morsey, J.D. ; Chew, W.C. ; Deutsch, A. ; Smith, H.H.
Author_Institution :
Dept. Electr. & Comput. Eng., Univ. of Illinois at Urbana-Champaign, Urbana, IL
Abstract :
A parallel LU decomposition algorithm is presented to take advantage of the sparse impedance matrix produced by the reduced-coupling method. This algorithm allows rapid simulation of very large chip and packaging problems. A representative example is shown for a wide, on-chip data-bus that required one million surface unknowns and the computational power of a 1024-node IBM BlueGene cluster with distributed memory.
Keywords :
distributed memory systems; impedance matrix; method of moments; sparse matrices; system buses; IBM BlueGene cluster; distributed memory; method-of-moments-based field solver; on-chip data-bus; packaging problems; product-level wide data-bus analysis; reduced-coupling technique; sparse impedance matrix; Clustering algorithms; Computational modeling; Concurrent computing; Data analysis; Distributed computing; Equations; Matrix decomposition; Packaging; Sparse matrices; Surface impedance;
Conference_Titel :
Electrical Performance of Electronic Packaging, 2007 IEEE
Conference_Location :
Atlanta, GA
Print_ISBN :
978-1-4244-0883-2
DOI :
10.1109/EPEP.2007.4387196
