Minimizing and Exploiting Leakage in VLSI Design (Repost)

Power consumption of VLSI (Very Large Scale Integrated) circuits has been growing at an alarmingly rapid rate. This increase in power consumption, coupled with the increasing demand for portable/hand-held electronics, has made power consumption a dominant concern in the design of VLSI circuits today. Traditionally, dynamic (switching) power has dominated the total power consumption of an IC. However, due to current scaling trends, leakage power has now become a major component of the total power consumption in VLSI circuits. Leakage power reduction is especially important in portable/hand-held electronics such as cell-phones and PDAs.
This book presents techniques aimed at reducing and exploiting leakage power in digital VLSI ICs. The first part of this book presents several approaches to reduce leakage in a circuit. The second part of this book shows readers how to turn the leakage problem into an opportunity, through the use of sub-threshold logic, with adaptive body bias to make the designs robust to variations. The third part of this book presents design and implementation details of a sub-threshold IC, using the ideas presented in the second part of this book.

Provides a variety of approaches to control and exploit leakage, including implicit approaches to find the leakage of all input vectors in a design, techniques to find the minimum leakage vector of a design (with and without circuit modification), ASIC approaches to drastically reduce leakage, and methods to find the optimal reverse bias voltage to maximally reduce leakage.
Presents a variation-tolerant, practical design methodology to implement sub-threshold logic using closed-loop adaptive body bias (ABB) and Network of PLA (NPLA) based design. In addition, asynchronous micropipelining techniques are presented, to substantially reclaim the speed penalty of sub-threshold design.
Validates the proposed ABB and NPLA sub-threshold design approach by implementing a BFSK transmitter design in the proposed design style. Test results from the fabricated IC are provided as well, indicating that a power improvement of 20X can be obtained for a 0.25um process (projected power improvements are 100X to 500X for 65nm processes).