Due to significant manufacturing process variations, the performance ofintegrated circuits (ICs) has become increasingly uncertain. Such uncertaintiesmust be carefully quantified with efficient stochastic circuit simulators. Thispaper discusses the recent advances of stochastic spectral circuit simulatorsbased on generalized polynomial chaos (gPC). Such techniques can handle bothGaussian and non-Gaussian random parameters, showing remarkable speedup overMonte Carlo for circuits with a small or medium number of parameters. We focuson the recently developed stochastic testing and the application ofconventional stochastic Galerkin and stochastic collocation schemes tononlinear circuit problems. The uncertainty quantification algorithms forstatic, transient and periodic steady-state simulations are presented alongwith some practical simulation results. Some open problems in this field arediscussed.
展开▼
