An Alternative Carry-save Arithmetic for New Generation Field Programmable Gate Arrays

Abstract

In this work, a double carry-save addition operation is proposed, which is efficiently synthesized for 6-input LUT-based eld programmable gate arrays (FPGAs). The proposed arithmetic operation is based on redundant number representation and provides carry propagation-free addition. Using the proposed arithmetic operation, a compact and fast multiply and accumulate unit is designed. To our knowledge, the proposed design provides the fastest multiply-add operation for 6-input LUT-based FPGA systems. A nite impulse response lter implementation is given to show the performance of the proposed structure. The proposed implementation provides a dramatic performance increase, which is at least 2 times faster than conventional binary multiply-add implementations.