基于PC集群的三维图形并行渲染性能分析  

Parallel rendering efficiency analysis of 3D graphics on PC cluster

作　　者：彭帅[1] 李冬梅[1] 李朝晖[1] 

机构地区：[1]中国传媒大学信息工程学院,北京100024

出　　处：《计算机应用》2010年第2期547-550,共4页journal of Computer Applications

基　　金：中国传媒大学211重点学科建设项目(21103040115)

摘　　要：研究基于PC集群的三维图形并行渲染性能问题,从网络性能、算法复杂度、并行分配机制等几方面分析了影响并行渲染性能的关键因素。在千兆以太网PC集群上进行了基于通用MPI和OpenGL的三维图形并行渲染仿真测试,给出了数据及分析结果,给出了合理构建并行三维图形渲染系统的建议,通过平衡图形算法复杂度和网络性能以达到最佳并行性能。In this paper, the efficiency of parallel 3D graphics rendering on general-purpose PC cluster was investigated. Three key factors affecting the efficiency of parallel rendering were analyzed, including network performance, algorithm complexity and parallel rendering mechanism. Parallel rendering simulations based on MPI and OpenGL were carried out on gigabit Ethernet PC cluster. The experimental results were analyzed and the advices for designing efficient.parallel rendering system on PC cluster were given.

关 键 词：并行渲染 OPENGL MPI 并行效率 

分 类 号：TP338.6[自动化与计算机技术—计算机系统结构]