Data-parallel volume rendering with adaptive volume subdivision

A data-parallel processing approach is promising for real-time volume rende ring because of the massive parallelism in volume rendering. In data-parall el volume rendering, local results processing elements(PEs) generate from a llocated subvolumes are integrated to form a final image. Generally, the in tegration causes an overhead unavoidable in data-parallel volume rendering due to communications among PEs. This paper proposes a data-parallel shear- warp volume rendering algorithm combined with an adaptive volume subdivisio n method to reduce the communication overhead and improve processing effici ency. We implement the parallel algorithm on a message-passing multiprocess or system for performance evaluation. The experimental results show that th e adaptive volume subdivision method can reduce the overhead and achieve hi gher efficiency compared with a conventional slab subdivision method.