Cost Effective Implementation of Fixed Point Adders for LUT based FPGAs using Technology Dependent Optimizations

Modern day field programmable gate arrays (FPGAs) have very huge and versatile logic resources resulting in the migration of their application domain from prototype designing to low and medium volume production designing. Unfortunately most of the work pertaining to FPGA implementations does not focus on the technology dependent optimizations that can implement a desired functionality with reduced cost. In this paper we consider the mapping of simple ripple carry fixed-point adders (RCA) on look-up table (LUT) based FPGAs. The objective is to transform the given RCA Boolean network into an optimized circuit netlist that can implement the desired functionality with minimum cost. We particularly focus on 6-input LUTs that are inherent in all the modern day FPGAs. Technology dependent optimizations are carried out to utilize this FPGA primitive efficiently and the result is compared against various adder designs. The implementation targets the XC5VLX30-3FF324 device from Xilinx Virtex-5 FPGA family. The cost of the circuit is expressed in terms of the resources utilized, critical path delay and the amount of on-chip power dissipated. Our implementation results show a reduction in resources usage by at least 50%; increase in speed by at least 10% and reduction in dynamic power dissipation by at least 30%. All this is achieved without any technology independent (architectural) modification