In the article the general going is considered near organization of calculable processes in the parallel system of 3D- stereo visualizations and reflections of procedure of 3D- stereo synthesis on architecture of GPU. Experimental research of realization of the modified algorithm of search of crossings is executed on GPU, as one of the most expense stages of RAYtRAYsing. The got results rotined the substantial increase of the productivity during многопоточной realization on architecture of GPU of the test system, which can be improved during realization on new micro architectures of GPU and optimizations of configuration of CUDA- of network.
3D- stereo visualization, image, synthesis, tracing of rays, accelerating techniques, architecture, GPU, realization, efficiency
"GPU-realyzatsyia parallelnoi vыchyslytelnoi systemы 3D-stereo vyzualyzatsyy s yspolzovanyem metoda trassyrovky luchei" [Visualization 3D-stereo parallel computer system GPU-realization with the use of rays tracing method],
Information Processing Systems,