Three-Dimensional Monte Carlo Device Simulation for Massively Parallel Architectures

H. Sheng, R. Guerrieri and Alberto L. Sangiovanni-Vincentelli

EECS Department
University of California, Berkeley
Technical Report No. UCB/ERL M95/53
July 1995

http://www2.eecs.berkeley.edu/Pubs/TechRpts/1995/ERL-95-53.pdf

The applicability of a massively parallel processing (MPP) paradigm to the simulation of charge transport in semiconductor devices through the Monte Carlo method is investigated. A unique mapping of Monte Carlo simulation to a data-parallel software model has been developed in which the problem is decoupled into multiple computational domains, thereby increasing the locality of computation. The target machine for the Monte Carlo simulator is the Connection Machine CM-2, a massively parallel SIMD architecture. Enhancement of convergence rates for Monte Carlo estimators are achieved through the development of methods for efficient carrier flight time generation and importance sampling. The computational efficiencies of the algorithms are analyzed, and applications of the simulator are shown.

BibTeX citation:

@techreport{Sheng:M95/53,
    Author = {Sheng, H. and Guerrieri, R. and Sangiovanni-Vincentelli, Alberto L.},
    Title = {Three-Dimensional Monte Carlo Device Simulation for Massively Parallel Architectures},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {1995},
    Month = {Jul},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/1995/2810.html},
    Number = {UCB/ERL M95/53},
    Abstract = {The applicability of a massively parallel processing (MPP) paradigm to the simulation of charge transport in semiconductor devices through the Monte Carlo method is investigated. A unique mapping of Monte Carlo simulation to a data-parallel software model has been developed in which the problem is decoupled into multiple computational domains, thereby increasing the locality of computation. The target machine for the Monte Carlo simulator is the Connection Machine CM-2, a massively parallel SIMD architecture. Enhancement of convergence rates for Monte Carlo estimators are achieved through the development of methods for efficient carrier flight time generation and importance sampling. The computational efficiencies of the algorithms are analyzed, and applications of the simulator are shown.}
}

EndNote citation:

%0 Report
%A Sheng, H.
%A Guerrieri, R.
%A Sangiovanni-Vincentelli, Alberto L.
%T Three-Dimensional Monte Carlo Device Simulation for Massively Parallel Architectures
%I EECS Department, University of California, Berkeley
%D 1995
%@ UCB/ERL M95/53
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/1995/2810.html
%F Sheng:M95/53