Wave computer core using fixed-point arithmetic

Abstract

In this paper, success on cellular nonlinear network emulator core that generates active waves such as autowaves and traveling waves is intended. This wave computer core has 4 × 4 parallel processing units. Cellular nonlinear network in this study has 16, 384 nodes arranged in 128 × 128 normal grid form. The evolution algorithm of the network is executed by the hardware implemented on a Xilinx XC2VP30-FF896 FPGA chip using fixed-point arithmetic. The FPGA-based platform has an on-line monitor output in order to observe active wave evolution and a host computer in order to program the network. Using fixed-point number format rather than floating-point number format has some advantages in the sense of speed and resource usage. The implementation in this study can be adapted to observe Doppler Effect on traveling waves and autowaves.