基于MMX技术的SIMD并行运算优化算法  

The SIMD Concurrent Operation Optimization Based on Bayesian net Works MMX Technique

作　　者：张剑[1] 

机构地区：[1]华中科技大学计算机科学与技术学院国家外存储系统专业实验室

出　　处：《传感技术学报》2005年第4期897-900,共4页Chinese Journal of Sensors and Actuators

基　　金：全国教育科学十五规划重点课题资助(AYA010034);国家自然科学基金资助(69735020)

摘　　要：将面向对象思想引入到SAD值计算的并行操作过程中,并从SIMD并行运算的角度出发,给出了改进的图像组织优化算法,通过对MMX优化后的编码器速度的测试结果知,在目前H.264/AVC的视频编码上,该编码器的编码速度有明显地提高,为实现窄带中的实时视频通信提供了保障。We give the image organization optimization algorithm of improvement from the image ot Single Instruction Multiple Data parallel operation, Such as the idea of oriented object to lead into Parallel operation process that calculated SAD value, through the test result of code device moving forecast for MultiMedia extension optimization velocity testing, we know that in current most complex video coding H.264/AVC, the implement of this algorithm can raise the coding speed of encoder obviously , and offer guarantee to realize the video communication of real time in narrow bandwidth.

关 键 词：编码器 矢量 并行运算 单指令多数据 

分 类 号：TN919.81[电子电信—通信与信息系统]