Fast arithmetic based on residue number system architectures

A new approach is described for transforming residue numbers into equivalent decimal numbers. The approach is based on the periodicity inherent in residue numbers and uses matrix techniques to achieve high-speed residue-to-decimal conversion. In comparison with existing conversion techniques, the proposed conversion method achieves a modest reduction in the number of arithmetic multiplications needed during the conversion process. Whether for achieving speed through parallelism in computer architectures or for enhancing the arithmetic reliability, residue number architectures are a promising approach for the future