This paper presents a divide-and-conquer ray-traced volume rendering algorithm and a parallel image compositing method, along with their implementation and performance on the Connection Machine CM-5, and networked workstations. This algorithm distributes both the data and the computations to individual processing units to achieve fast, high-quality rendering of high-resolution data. The volume data, once distributed, is left intact. The processing nodes perform local raytracing of their subvolume concurrently. No communication between processing units is needed during this locally ray-tracing process. A subimage is generated by each processing unit and the #nal image is obtained by compositing subimages in the proper order, which can be determined a priori. Test results on both the CM-5 and a group of networked workstations demonstrate the practicality of our rendering algorithm and compositing method. y This researchwas supported in part by the National Aeronautics and Space Administration under NASA contract NAS1-19480 while the author was in residence at the Institute for Computer Application in Science and Engineering #ICASE#, NASA Langley Research Center, Hampton, VA 23681-0001. i 1

In computer graphics, rendering is the process by which an abstract description of a scene is converted to an image. When the scene is complex, or when high-quality images or high frame rates are required, the rendering process becomes computationally demanding. To provide the necessary levels of performance, parallel computing techniques must be brought to bear. Although parallelism has been exploited in computer graphics since the early days of the field, its initial use was primarily in specialized applications. The VLSI revolution of the late 1970Õs and the advent of scalable parallel computers during the late 1980Õs changed this situation. Today, parallel hardware is routinely used in graphics workstations, and numerous software-based rendering systems have been developed for general-purpose parallel architectures. This article provides a broad introduction to the subject of parallel rendering, encompassing both hardware and software systems. The focus is on the underlying concepts and the issues which arise in the design of parallel rendering algorithms and systems. We examine the different types of parallelism and how they can be applied in rendering applications. Concepts from parallel computing, such as data decomposition, task granularity, scalability, and load balancing, are considered in relation to the rendering

A parallel solution to the visualisation of high resolution vol- ume data is presented. Based on the ray tracing (RT) visu- alization technique, the system works on a distributed memory MIMD architecture. A hybrid strategy to ray tracing parallelization is applied, using ray dataflow within an image partition approach. This strategy allows the flexible and effective management of huge dataset on architectures with limited local memory. The dataset is distributed over the nodes using a slice-partitioning technique. The simple data partition chosen implies a straighforward communications pattern of the visualization processes and this improves both software design and eJciency, while providing deadlock prevention. The partitioning technique used and the network interconnection topology allow for the efficient implementation of a statical load balancing technique through pre-rendering of a low resolution image. Details related to the practical issues involved in the parallelization of volumetric RT are discussed, with particular reference to deadlock and termi- nation issues.

Rendering volumes represented as a 3D grid of voxels requires an overwhelming amount of processing power. In this paper we investigate efficient techniques for rendering semi-transparent volumes on vector and parallel processors. Parallelism inherent in a regular grid is obtained by decomposing the volume into geometric primitives called beams, slices and slabs of voxels. By using the adjacency properties of voxels in beams and slices, efficient incremental transformation schemes are developed. The slab decomposition of the volume allows the implementation of an efficient parallel feed-forward renderer which includes the splatting technique for image reconstruction and a back-to-front method for creating images. We report the implementation of this feed-forward volume renderer on a hierarchical shared memory machine with individual pipelined processors. 1: Introduction Representation by spatial-occupancy enumeration methods allows a simple yet versatile method for the generation and ...

Volume rendering algorithms visualize sampled three dimensional data. A variety of applications create sampled data, including medical imaging, simulations, animation, and remote sensing. Researchers have sought to speed up volume rendering because of the high run time and wide application. Our algorithm uses permutation warping to achieve linear speedup on data parallel machines. This new algorithm calculates higher quality images than previous distributed approaches, and also provides more view angle freedom. We present permutation warping results on the SIMD MasPar MP-1. The efficiency results from nonconflicting communication. The communication remains efficient with arbitrary view directions, larger data sets, larger parallel machines, and high order filters. We show constant run time versus view angle, tunable filter quality, and efficient memory implementation. 1 Introduction Volume rendering [4] is memory and compute bound. Researchers have used parallelism to speedup transpa...

We describeatechnique for achieving fast volume ray casting on parallel machines, using a load balancing scheme and an e#cient pipelined approach to compositing. We propose a new model for measuring the amount of work one needs to perform in order to render a given volume, and use this model to obtain a better load balancing scheme for distributed memory machines. We also discuss in detail the design tradeo #s of our technique. In order to validate our model we have implemented it on the Intel iPSC#860 and the Intel Paragon, and conducted a detailedperformance analysis. 1 Introduction As researchers and engineers use volume rendering to study complex physical and abstract structures they need a coherent, powerful, easy to use visualization tool, that lets them interactively change all the necessary parameters. Unfortunately,even with the latest volume rendering acceleration techniques running on top-of-the-line workstations, it still takes a few seconds to a few minutes to volume ren...

Volume rendering is a key technique in scientific visualization that lends itself to significant exploitable parallelism. The high computational demands of real-time volume rendering and continued technological advances in the area of VLSI give impetus to the development of special-purpose volume rendering architectures. This paper presents and characterizes three recently developed volume rendering engines which are based on the ray-casting method. A taxonomy of the algorithmic variants of ray-casting and details of each ray-casting architecture are discussed. The paper then compares the machine features and provides an outlook on future developments in the area of volume rendering hardware.

this paper we consider a general scheme for parallel volume rendering, described in Section 2. We describe an algorithm in this paper that has an extremely efficient implementation on distributed memory MIMD architectures and is suitable for hardware implementation based on the image composition architectures [15]. This data parallel scheme assigns a portion of the volume to each processor which renders it with any one of the above mentioned rendering algorithms. The resulting images from all processors are then combined (composited) in visibility order to form the final image. As viewing and shading parameters change, 3D voxel data is not communicated between processors. Communication involves only 2D partial images. In Section 3 we report on specific rendering algorithms that can be executed in each processor to generate the image of the local subvolume. These algorithms take advantage of vector processing or pipelining capabilities available on some parallel machines (e.g., CRAY Y-MP, Intel Paragon, IBM Power Visualization System). Section 4 contains performance analysis and provides results of numerical experiments examining various aspects of the proposed algorithm and its variations, while Section 5 contains concluding remarks and future plans. 2. Volume Rendering by Parallel Image Composition

ix Acknowledgements x Acknowledgements : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : x Publication History : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : x 1. Introduction 1 1.1 Introduction to Direct Volume Rendering : : : : : : : : : : : : : : : : : : : 2 1.1.1 Volumetric Grids : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 3 1.1.2 Image-space Rendering Algorithms : : : : : : : : : : : : : : : : : : : 4 1.1.3 Object-space Rendering Algorithms : : : : : : : : : : : : : : : : : : 5 1.1.4 Shear Transformations : : : : : : : : : : : : : : : : : : : : : : : : : : 7 1.1.5 Complexity : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 7 1.2 Motivation for Parallel Direct Volume Rendering : : : : : : : : : : : : : : : 8 1.2.1 Scalability Is Important : : : : : : : : : : : : : : : : : : : : : : : : : 8 1.3 Context for Use : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 9 1.3.1 Distributed Graphical Us...

We present the CellFlow method for object space subdivision that exploits frame coherency to implement a look-ahead scheme of object dataflow. The implementation of this scheme exploits the communication features of modern scalable multicomputers to achieve near linear speedup by means of latency hiding. We demonstrate the performance of our approach in the field of volume rendering by implementing incremental rotation of the volumetric object about its center. The simplicity of the algorithm, its optimal embedding in popular network topologies, and minimal congestionfree communication among processors are its main advantages. Results are shown for implementation on the Cray T3D, a distributed memory 3D torus architecture. Computation and communication load balancing issues among processors are also addressed. 1. Introduction Medical data obtained from MRI (magnetic resonance imaging) and CT-scanners (computed tomography) act as good sources for volume visualization applications. This...