Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy (PhD)

Degree Program

Electrical and Computer Engineering

Year Degree Awarded

Spring 2015

First Advisor

Wayne Burleson

Subject Categories

Digital Circuits | Hardware Systems | VLSI and Circuits, Embedded and Hardware Systems

Abstract

Advanced CMOS technologies have enabled high density designs at the cost of complex fabrication process. Variation in oxide thickness and Random Dopant Fluctuation (RDF) lead to variation in transistor threshold voltage Vth. Current photo-lithography process used for printing decreasing critical dimensions result in variation in transistor channel length and width. A related challenge in nanometer CMOS is that of on-chip random noise. With decreasing threshold voltage and operating voltage; and increasing operating temperature, CMOS devices are more sensitive to random on-chip noise in advanced technologies.

In this thesis, we explore novel circuit techniques to manage the impact of process variation in nanometer CMOS technologies. We also analyze the impact of on-chip noise on CMOS circuits and propose techniques to leverage or manage impact of noise based on the application. True Random Number Generator (TRNG) is an interesting cryptographic primitive that leverages on-chip noise to generate random bits; however, it is highly sensitive to process variation. We explore novel metastability circuits to alleviate the impact of variations and at the same time leverage on-chip noise sources like Random Thermal Noise and Random Telegraph Noise (RTN) to generate high quality random bits. We develop stochastic models for metastability based TRNG circuits to analyze the impact of variation and noise. The stochastic models are used to analyze and compare low power, energy efficient and lightweight post-processing techniques targeted to low power applications like System on Chip (SoC) and RFID. We also propose variation aware circuit calibration techniques to increase reliability. We extended this technique to a more generic application of designing Post-Si Tunable (PST) clock buffers to increase parametric yield in the presence of process variation. Apart from one time variation due to fabrication process, transistors undergo constant change in threshold voltage due to aging/wear-out effects and RTN. Process variation affects conventional sensors and introduces inaccuracies during measurement. We present a lightweight wear-out sensor that is tolerant to process variation and provides a fine grained wear-out sensing. A similar circuit is designed to sense fluctuation in transistor threshold voltage due to RTN. Although thermal noise and RTN are leveraged in applications like TRNG, they affect the stability of sensitive circuits like Static Random Access Memory (SRAM). We analyze the impact of on-chip noise on Bit Error Rate (BER) and post-Si test coverage of SRAM cells.

Share

COinS