A generic method for variability analysis of nanoscale circuits

We present a generic method for analyzing the effect of process variability in nanoscale circuits. The proposed framework uses kernel and a generic tail probability estimator to eliminate the need for a-priori density choice for the nature of circuit variation. This allows capturing the true nature of the circuit variation from a few random samples of its observed responses. The data-driven, non-parametric, and generic nature of the proposed method can greatly facilitate the technology-circuit co-design.