作为高性能计算领域的新兴应用,渲染有着计算密集和数据密集的应用特征,如何提高渲染农场的效率进行大规模快速渲染,近年来逐渐成为研究的热点问题。对渲染农场的国内外应用现状作了介绍,阐述了渲染农场的组织结构和工作流程。针对渲染农场在优化作业调度算法、改善负载均衡策略以及完善渲染管理平台等三个方面的研究,进行了全面的分析和比较。探讨和总结了在云计算背景下渲染农场在渲染作业调度策略、异构集群渲染任务分发算法、基于云模式的渲染农场新架构和负载均衡策略这四个方面中有待进一步研究的问题。
As a newly-developing application in the field of high performance computing, render has the computing intensive and data intensive application characteristics. It has become the hot issue gradually how the efficiency of rendering farm for large-scale and fast rendering is improved in recent years. Firstly, this paper introduced the application status for render farm and then expounded the architecture and workflow of render farm. Secondly, it analyzed and compared comprehensively the re- search on optimization of job scheduling algorithm, improvement of load balancing strategy and perfection of render manage- ment platform. Finally, under the background of cloud computing it discussed and summarized the research problems to be re- solved on rendering job scheduling method, rendering task distribution algorithm in heterogeneous cluster, the new render farm architecture based on cloud schema and load balancing strategy.