RAY CASTING ON SHARED-MEMORY ARCHITECTURES - MEMORY-HIERARCHY CONSIDERATIONS IN VOLUME RENDERING

The authors provide a detailed analysis of the memory-hierarchy effect s in shared-memory architectures of one method of volume rendering-ray casting. They studied two parallel-partitioning and dynamic load-bala ncing algorithms-one object partition and one image partition-explorin g trade-offs between their memory-hierarchy performance and the algori thmic optimizations they allow. Their resulting implementations (along with careful tuning of the ray-advancement kernel for Silicon Graphic s' R8000) yield extremely high performance. For a 1-Gbyte female human -body data set, they attain an average frame rate of 1.0 frame per sec ond, at a resolution of 400 pixels x 300 pixels, on a 16-processor Sil icon Graphics Power Challenge. This is faster than the literature has previously reported for a data set this large. They have also extended their methods to a cluster of such machines. Using eight Silicon Grap hics Power Challenge machines with a total of 64 processors, they atta in average frame rates up to 10 frames per second on a 357-Mbyte male human-body data set for a sequence of frames generated by interactive user control.