Combinatorial Logic Synthesis using Technology Directed Decomposition
K Santeppa, K. A. Neelakantan, Vaidyeswaran Rajaraman
Abstract
K Santeppa, K. A. Neelakantan, Vaidyeswaran Rajaraman
Abstract
The aim of logic synthesis is to produce circuits which satisfy the given boolean function while meeting timing constraints and requiring the minimum silicon area. Logic synthesis involves two steps namely logic decomposition and technology mapping. Existing methods treat the two as separate operation. The traditional approach is to minimize the number of literals without considering the target technology during the decomposition phase. The decomposed expressions are then mapped on to the target technology to optimize the area. Timing optimization is carried out subsequently. A new approach which treats logic decomposition and technology maping as a single operation is presented. The logic decomposition is based on the parameters of the target technology. The area and timing optimization is carried out during logic decomposition phase itself. Results using MCNC circuits are presented to show that this method produces circuits which are 38% faster while requiring 14% increase in area.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The aim of logic synthesis is to produce circuits which satisfy the given boolean function while meeting timing constraints and requiring the minimum silicon area. Logic synthesis involves two steps namely logic decomposition and technology mapping. Existing methods treat the two as separate operation. The traditional approach is to minimize the number of literals without considering the target technology during the decomposition phase. The decomposed expressions are then mapped on to the target technology to optimize the area. Timing optimization is carried out subsequently. A new approach which treats logic decomposition and technology maping as a single operation is presented. The logic decomposition is based on the parameters of the target technology. The area and timing optimization is carried out during logic decomposition phase itself. Results using MCNC circuits are presented to show that this method produces circuits which are 38% faster while requiring 14% increase in area.
Key concepts: Logic optimization, Decomposition, Logic synthesis, Sequential logic, Logic family, Logic gate, Functional decomposition, Computer science