Non-parametric statistical static timing analysis: An SSTA framework for arbitrary distribution

We present a new statistical STA framework based on Monte Carlo analysis that can deal with arbitrary statistical distribution and delay models. Order statistics (non-parametrics) is consistently adopted by which the timing analysis and criticality calculation become distribution-independent. To make Monte Carlo process computationally practical, delays are handled as vectors so that iterations are eliminated. The vector dimension or required number of Monte Carlo iterations which guarantees no timing violation at any user- specified probability is analytically determined. A path criticality metric using order statistics is also defined. Experimental results using various delay models show the validity and usefulness of our proposed algorithm.