This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: BDD-B ASED L OGIC O PTIMIZATION S YSTEM Congguang Yang Maciej Ciesielski Feburary 2000 TR-CSE-00-1 cyang,[email protected] Department of Electrical and Computer Engineering University of Massachusetts Amherst, MA 01003 YANG AND CIESIELSKI: BDD-BASED LOGIC OPTIMIZATION SYSTEM 2 BDD-B ASED L OGIC O PTIMIZATION S YSTEM Congguang Yang Maciej Ciesielski cyang,ciesiel @ecs.umass.edu Department of Electrical & Computer Engineering University of Massachusetts I. I NTRODUCTION Logic synthesis plays a central role in the design automation of VLSI circuits. Software tools for logic synthesis are one of the most important tools ever developed in the area of Computer-Aided Design (CAD). With the help of those tools, a designer is freed from tedious and error-prone low-level circuit design, and can focus on architectural and algorithmic level issues. Logic synthesis is composed of three main steps. First, a circuit described in high-level language (hardware description languages, such as VHDL or Verilog) is transformed into a Boolean network. Then, the Boolean network is optimized using logic optimization tools. Finally, the optimized Boolean network is mapped to a library of logic cells. The entire process is directed in such a way as to optimize certain design objectives (such as delay, area, power, etc) and meet users’ specifications and constraints. Among these three steps, logic optimization is the most important. Because the quality of final synthesis results is mainly determined by it. As a result, intensive research has been done in this area. A. Traditional Multi-Level Logic Optimization The main theme in multi-level logic optimization is factorization . In a typical logic synthesis environment, a Boolean function is initially represented as a sum-of- product (SOP) or cube form. This form is transformed by factoring out common algebraic or Boolean expressions. In an algebraic factorization, logic functions are treated as polynomials, in which rules of Boolean algebra are not applied. Boolean factorizations, based on Boolean division, apply Boolean algebra rules, hence can produce better results in terms of the resulting logic complexity (number of terms, literals, etc). Traditional logic optimization methodology, based on algebraic factorization for Boolean networks , , has gained tremendous success in logic optimization and emerged as the dominant method. However, while near optimal results can be obtained for those Boolean functions which can be represented with AND/OR expressions, results are far from satisfactory for functions which can be compactly represented as a combination of AND/OR and XOR expressions. This work has been supported by a grant from NSF under contract No....
View Full Document
- Fall '09
- Logic, Boolean Algebra, BDDs, Congguang Yang