A prototype implementation of OpenGL SC on a massively-parallel architecture

OpenGL SC (OpenGL for Safety Critical) is the safety critical variation of the widely-used OpenGL standard, and designed to meet the needs of safety critical markets for avionics, industrial, military, medical and automotive applications. We naturally need a cost-effective way of implementing OpenGL SC. For more flexibility on the graphics processing, we need a full-software 3D graphics pipeline implementation. From the programmer´s point of view, this kind of software implementation is essential to access the internal features of the traditional graphics pipeline. Nowadays, the massively-parallel rasterization engines, which are based on the general-purpose GPU programming features, showed that those full software processing is possible. In this paper, we present a prototype implementation of OpenGL SC on a massively parallel architecture. We choose CUDA as the target parallel architecture. Our prototype implementation is verified by several demonstration programs. Verifications on the real-time systems will be followed. After fully verifying our implementation on the massively parallel architecture, we will make it commercially available.