Communication Costs of Parallel Volume Rendering of Large Images on a Network Of Workstations

Direct Volume Rendering is a popular technique for visu alization of D datasets that o ers many advantages over other algo rithms but requires important computing power E cient parallelization is therefore an essential need Taking advantage of an existing LAN in stead of using a dedicated parallel architecture can be a cost e ective solution However this requires special attention in algorithm design to the speci c aspects of a LAN as a parallel computing system long la tency slow throughput Besides due to their large volume up to GB D datasets as those produced by modern medical scanners can not be made entirely resident into main memory We present in this paper the study of an algorithm for direct volume rendering of large images on a local network of workstations which does not assume the presence of the whole dataset in main memory A method for handling large dis tributed D image datasets is proposed and three parallelization strate gies are compared based on communication performance measurements of a LAN