Design Of Floating-Point Calculator Based On FPGA

With the growth demand in the field of electronics industry application, in order to achieve a higher degree of complexity of the digital electronic system, the demands for data-processing capabilities are also very high. Fixed-point calculation have been difficult to meet the requirements of high-performance digital systems, by comparison, floating-point numbers as opposed to fixed-point, has the advantages of wider statement range and higher valid precision. Therefore, it had widely used in aerospace industry, remote sensing, robot technology, as well as index calculation and signal processing areas. The requirements of the floating-point calculation are mainly reflected in two aspects: Firstly, the speed, that is how fast and effective completion of floating-point calculation; Secondly, the precision, means that the floating-point calculation can provide how many valid numbers.The improvement of cost-effective as well as the emergence of programmable logic devices made the methods of traditional digital electronic system design be changed. FPGA (Field Programmable Gate Array) made developers design chip to implement the functions of electronic system, both the selection of components and circuit board design were carried out in the process of chip design. FPGA can implement complex functions of timing and composite logic circuit, suitable for high-speed, high density logic design field with complex algorithm of the logic unit and signal processing unit, such as calculator, digital filter and 2-d convolver.In view of the characteristics of FPGA technology and the extensive use of floating-point calculation, this paper combines floating-point calculation of FPGA and practical application to design a floating-point calculator. The main purpose is to implement the basic calculating functions of floating-point numbers via programming of VHDL language, such as addition, subtraction, multiplication, division and extraction.(1) The overall framework of system design and realization of each module which contain selection of chip, timing and control between modules, detailed principle and design process of algorithm for each module were all described;(2) The basic calculating functions of floating-point numbers, such as addition, subtraction, multiplication, division and extraction were implemented with VHDL;(3) In the circumstance of Xilinx ISE, the development and functional simulation for main modules of system were accomplished, and then floating-point calculation based on FPGA was confirmed.