GOAL-DRIVEN SEARCH OF A STOCHASTIC PROCESS USING REDUCED SETS OF SIMULATION POINTS

Various techniques for improving performance of a goal-seeking search of a computer-simulated stochastic process are disclosed. One such technique may include generating an N-point Monte Carlo simulation of a stochastic model, such as a model representative of a digital electronic circuit, and selecting a subset of M points from the N-point Monte Carlo simulation, where M is less than N. The technique may further include searching the subset of M points to identify a target value, wherein said searching comprises generating one or more M-point Monte Carlo simulations of the stochastic model; and checking the target value, wherein said checking comprises generating an additional N-point Monte Carlo simulation of the stochastic model dependent on results of searching the subset of M points.