ENEE 759B: Advances in Low-Power Design Methodologies

Course Goals:

Driven by increased levels of device integration and complexity, together with higher device speed, power dissipation has become a crucial design concern, limiting the number of devices that can be put on a chip and dramatically affecting the packaging and cooling costs associated with ASICs. Power dissipation is even a bigger concern for the class of battery-powered personal computing devices and wireless communication systems. To make low-power design a reality, this course focuses on the challenges in low-power design of CMOS digital circuits and provides an overview of state-of-the-art techniques for power estimation and optimization at different levels of abstraction, from circuit level to architectural level. An emphasis will be placed on analytic techniques based on probabilistic models which have a large area of application in low-power systems. In particular, the applications of finite order Markov chains in static power estimation, state assignment for low-power, and sequence compaction will be discussed in detail. Finally, the course will also examine the applications of some information-theoretic concepts in power estimation and optimization at RT- and behavioral-level of abstraction.

Course Prerequisite(s):

ENEE 644, ENEE 640 or permission of the instructor

Topics Prerequisite(s):

Basic VLSI design knowledge, logic and high-level synthesis and optimization, basic probability theory concepts.

Textbook(s):

None

Reference(s):

N. Weste and K. Eshraghian, `Principles of CMOS VLSI Design: A Systems Perspective', 2nd Edition, Addison-Wesley Publishing Co., 1994.
J. M. Rabaey and M. Pedram (eds.), `Low Power Design Methodologies,' Kluwer Academic Publishers, 1996.
M. Pedram, `Power Minimization in IC Design: Principles and Applications', in ACM Transactions on Design Automation of Electronic Systems, vol. 1, no. 1, pp.1-54, Jan. 1996.

Core Topics:

  • Sources of power dissipation in CMOS circuits
  • Static power estimation techniques (including probabilistic simulation, probabilistic modeling of dependencies, macro-modeling)
  • Dynamic power estimation techniques (including sampling techniques and sequence compaction)
  • Power optimization at circuit level (including transistor resizing)
  • Power optimization at logic level (including multilevel and two level optimization)
  • Power optimization at RT- and behavioral-level (including module assignment, register sharing for low power, multiple voltage assignment)

Optional Topics:

  • Energy recovery CMOS
  • Low power microprocessor design
  • Low power memory design

Information for Fall 98 Semester:

Instructor: Diana Marculescu Web page:
Phone: 405-6703 e-mail: dianam@eng.umd.edu
Office Hours: By appointment Office: AVW 1419
Time of Course: MW 9:30-10:45am Room: JMP 2202

Course Structure:

Each part of the course involves extensive reading of research papers on the subject. In addition to completing all reading assignments from the course pack, the course requirement consists of homework assignments, a final project and a class presentation of the final project. Student performance on each of these requirements will contribute to the student's final grade.

Grading Method:

The student performance will be evaluated using using the following breakdown:
Homeworks: 40%
Final project: 40%
Project presentation: 20%
The homework assignments will be based on the material discussed in the class as well as some research papers included in the course pack. Some of the homework assignments may imply programming and/or evaluations of different designs using commercial software tools. Finally, each student will be required to undertake a final project and give a short presentation in the class. The instructor will define several topics for the final project but students are encouraged to propose new topics and discuss them with the instructor. The projects will be individually tailored but students will also be allowed to work in small teams on more complex projects. During the class presentation of the projects, each student will have the opportunity to present its own contribution to the project and suggest possible directions for future work.

Last Updated:

July 9, 1998, Diana Marculescu

khodary@eng.umd.edu