Research On Parallel Depth Buffer Algorithm For Ternary Optical Computer Based On Depth Blanking Problem

In recent years,with the development of optical computing technology,the technology and application of ternary optical computer have become increasingly mature.Its high computing power,numerous data-bits,and the characteristics of bit by bit reconfigurable processor make it have great advantages in solving the problem of high data.In this paper,we proposed the method of pixel computaton of ternary optical computer and parallel depth buffer algorithm.The depth buffer algorithm is a pixel blanking method on the hardware level,which can realize collision detection and interference calculation of objects in space,and calculate the depth value of graphics in 3D scenes to help build the view model.The traditional depth buffer algorithm can not be separated from pixel by pixel operation,and has the disadvantages of slow speed,low computational efficiency and large space consumption.The ternary optical computer has a large number of data-bits,which can be allocated by bits.Therefore,a large number of computing units can be reconstructed on the optical processor to achieve parallel pixel computing.This paper first introduces the structure of ternary optical computer and double rotator structure ternary optical processor,then discusses the application and implementation of depth buffer algorithm in ternary optical computer,proposes a scheme of processor reconstruction and parallelism for this algorithm,and proves the correctness of this scheme through examples and experiments on existing platforms,and analyzes the feasibility of hardware resources,clock cycles and overlay scheme.Compared with traditional algorithms,the parallel depth buffer algorithm based on ternary optical computer has greater advantages in resources,time,efficiency,etc.It has certain application significance,and also provides methods and ideas for ternary optical computer platform to perform computing intensive tasks.The main research of the paper include:1.A scheme of ternary optical computer processing graphic computing is proposed,that is,the processor reconstructs a large number of computing units to realize pixel computing in a parallel way.The scheme refers to the architecture design of the graphics processing unit and fills the vacancy in the field of ternary optical computer.2.Using the continuity of the image,the image is divided into grids and blocks.It is calculated in the form of blocks.The processor bit of the ternary optical computer is divided into channels corresponding to the image pixels.Through the cooperative work of electronic computer and ternary optical computer,a platform with high computing power is built,and the work is allocated to special hardware for processing,so as to reduce the workload of the central processor of the electronic computer and improve the overall parallelism of the system.3.A unified component for depth buffer algorithm and interference detection is proposed,and the stacking component for output results is optimized.A multi data-bit stacking scheme is designed to support numerical comparison of any length and unified output of optical components.