Relaxation of time-dependent NBTI variability and separation from RTN

Abstract

NBTI and RTN time-dependent variability is described from a defect-centric perspective. It is shown that NBTI induced threshold voltage shift (ΔV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TH</sub> ) distribution is governed by a compound Exponential-Poisson process. Using the memoryless properties of Poisson statistics, it is shown that not only the stressed but also the relaxed fraction follows the Exponential-Poisson distribution. Moreover, the relaxed fraction is shown to be independent from the final remaining ΔV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TH</sub> . Using the same statistical model, the separation of RTN induced ΔV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TH</sub> from NBTI is shown which allows a more accurate estimation of the characteristic distribution parameters. Finally, it is shown that RTN is a wide-sense stationary noise source of which the autocorrelation can be extracted from analysis of NBTI relaxation traces.