VOLUME RENDERING DEVICE, VOLUME GRAPHIC INTEGRATED CIRCUIT, AND VOLUME RENDERING INTEGRATING CIRCUIT

摘要

PROBLEM TO BE SOLVED: To more improve cost performance and to realize high-performance, high-speed processing by allowing one of plural pipelines to send data to only one adjacent pipeline. SOLUTION: A VRC 202 is equipped with a processing element 210 having four parallel rendering pipelines 212 and a render controller 214, and the processing element 210 obtains voxel data from a voxel memory through a voxel memory interface 216 and provides them for a pixel memory through a pixel memory interface 218. The four pipelines 212-0 to 212-3 operate in parallel in an X direction and voxels are outputted in their scanning order to pipelines 212-0 to 212-3 in four voxel groups. Intermediate values needed for X-directional calculation are sent from respective pipelines (e.g. 212-0) to adjacent pipelines (e.g. 212-1) at proper time.