Post-silicon performance modeling and tuning of analog/mixed-signal circuits via Bayesian Model Fusion

Post-silicon tuning has recently emerged as an important technique to combat large-scale uncertainties (e.g., process variation, device modeling errors, etc) for today's nanoscale circuits. This talk presents a novel Bayesian Model Fusion (BMF) technique for efficient post-silicon performance modeling and tuning of analog and mixed-signal (AMS) circuits. The key idea is to borrow the simulation or measurement data from an early stage (e.g., pre-silicon) to accurately build AMS performance models at a late stage (e.g., post-silicon). The post-silicon models are then used to facilitate efficient tuning of AMS circuits. A circuit example designed in a commercial 32 nm CMOS process is used to demonstrate the efficacy of the proposed post-silicon performance modeling and tuning methodology based on BMF.