ABSTRACT

In the past, the major concern of VLSI designers were performance, area, reliability, and cost, with power being only a secondary issue. However, in recent years this has changed and power, area, and speed have become equally important. There are many reasons for this new trend. Primarily, the rapid advancement in semiconductor technology in the last decade has made possible the integration of a large number of digital CMOS circuits on a single chip. Moreover, the desirability of portable operations of these circuits has necessitated the development of low power technology. Portable applications could be anywhere from desk-tops to audio/video based multimedia products to personal digital assistants and personal communicators. These systems demand both complex functionality and low power at the same time, thereby making their design challenging. The power consumption of portable circuits has a direct bearing on the life-time of the batteries. For example, a portable multimedia terminal designed using off-the-shelf components (not optimized for low power), could consume about 30–40 watts of power. If this system were to use the state-of-the-art nickel-metal-hydride battery technology [1], it would require 4.5–6 kilograms of batteries for 10 hours of operation. Therefore, this would mean that portable systems will experience either heavy battery packs or a very short battery life. Reduction in power consumption also plays an important role for producers of non-portable systems. The state of the art microprocessors optimized for performance consume around 20–30 watts of power for operating frequencies of 150–200 MHz. With rapid advancement in technology, the speeds could reach 500–600 MHz with extraordinarily high power consumption values. This would mean that the packaging cost for such devices would be very high and expensive cooling and packaging strategies would be required. Therefore, reduction in power consumption could greatly cut cooling costs. Finally, the issue of reliability is also a 742major concern for consumer system designers. Systems which consume more power often run hot and acerbate failure mechanisms. In fact the failure rate increases rapidly for a small increase in operating temperature. Therefore, the maximum power consumption of the system is a crucial design factor as it could have an impact on the system cost, battery type, heat sinks, etc. Therefore, reduction in peak power is also an important issue. It is clear that the motivations for reduction in power consumption vary from application to application. In portable applications such as cellular phones and personal digital assistants, the goal is to keep the battery lifetime and weight reasonable. For high performance portable computers such as laptops, the goal is to reduce the power dissipation of the electronics portion of the system. Finally, for non-portable systems such as workstations and communication systems the goal is to reduce packaging, cooling cost and ensure long-term reliability.