基于PVM的OpenGL的并行实现

IMPLEMENTATION OF OPENGL ON PARALLEL VIRTUAL MACHINE

本文在以微机为节点的并行虚拟机（PVM）上实现了三维图形工业标准OpenGL，具有实用价值．文中提出了新的图像平面分割算法和图像区域的分块绘制与回传方法，使得各处理机之间的数据交换量较小且分布在整个绘制过程中．该方法特别适合于通信开销大的网络分布式计算环境．利用帧间一致性和线性预测，使得任务的分配与绘制并行完成，避免了等待．本文的研究结果可直接应用于建立基于微机网络的交互式实时三维图形系统．

A 3D polygon rendering system conforming to the specification of OpenGL has beenimplemented on PVM (Parallel Virtual Machine). The originality of our system lies in two. First,The system is targeted to low speed network of low cost serial computers, instead of dedicatedhardware; second, our polygon rendering system conforms to the specification of OpenGL which isa system independent, industrial standard interface for 3D graphics programming. In our parallelrendering algorithm, image space is decomposed to exactly number of regions equal to the numbero f processors which reduces the volume of communication and the polygons needed to be renderedmore than once on separate renderers. We also propose a transmission scheme that distributes thepassing of data throughout the whole process of rendering. Taking advantage of frame-to-framecoherence, load balancing is achieved by decomposing the image space unequally so that the workloadsin to each space region are proportional to the processing power of corresponding renders.distribution of workload and processing power that is volatile during rendering are obtained with linearprediction from statistics of previous frames. The system is practical and scalable for moderate numberof processors.