Book Description

With the increase of density and complexity of high-performance integrated circuits and systems, including many-core chips and system-on-chip (SoC), it is becoming difficult to meet the power delivery and regulation requirements with off-chip regulators. The off-chip regulators become a less attractive choice because of the higher overheads and complexity imposed by the additional wires, pins, and pads. The increased I2R loss makes it challenging to maintain the integrity of different voltage domains under a lower supply voltage environment in the smaller technology nodes. Fully integrated on-chip voltage regulators have proven to be an effective solution to mitigate power delivery and integrity issues. Two types of regulators are considered as most promising for on-chip implementation: (i) the low-drop-out (LDO) regulator and (ii) the switched-capacitor (SC)regulator. The first part of our research mainly focused on the LDO regulator. Inspired by the recent surge of interest for cap-less voltage regulators, we presented two fully on-chip external capacitor-less low-dropout voltage regulator design. The second part of this proposal explores the complexity of designing each block of the regulator/analog circuit and proposed a design methodology for analog circuit synthesis using simulation and learning-based approach. As the complexity is increasing day-by-day in an analog circuit, hierarchical flow mostly uses for design automation. In this work, we focused mainly on Circuit-level, one of the significant steps in the flow. We presented a novel, efficient circuit synthesis flow based on simulation and learning-based optimization methods. The proposed methodology has two phases: the learning phase and the evaluation phase. Random forest, a supervised learning is used to reduce the sample points in the design space and iteration number during the learning phase. Additionally, symmetric constraints are used further to reduce the iteration number during the sizing process. We introduced a three-step circuit synthesis flow to automate the analog circuit design. We used H-spice as a simulation tool during the evaluation phase of the proposed methodology. The three most common analog circuits are chosen: single-stage differential amplifier, operational transconductance amplifier, and two-stage differential amplifier to verify the algorithm. The tool is developed in Python, and the technology we used is 0.6um. We also verified the optimized result in Cadence Virtuoso.