OptiX

Abstract

The NVIDIA® OptiX™ ray tracing engine is a programmable system designed for NVIDIA GPUs and other highly parallel architectures. The OptiX engine builds on the key observation that most ray tracing algorithms can be implemented using a small set of programmable operations. Consequently, the core of OptiX is a domain-specific just-in-time compiler that generates custom ray tracing kernels by combining user-supplied programs for ray generation, material shading, object intersection, and scene traversal. This enables the implementation of a highly diverse set of ray tracing-based algorithms and applications, including interactive rendering, offline rendering, collision detection systems, artificial intelligence queries, and scientific simulations such as sound propagation. OptiX achieves high performance through a compact object model and application of several ray tracing-specific compiler optimizations. For ease of use it exposes a single-ray programming model with full support for recursion and a dynamic dispatch mechanism similar to virtual function calls.

GPU computing in medical physics: A review

“…48 Complex tissue geometries can also be described using triangular meshes, but these methods are benefiting only modestly from GPU acceleration 49 because efficient GPU calculation of the intersection of a ray with a triangular mesh remains a challenging problem and an intense focus of research in computer graphics. 50 Simpler dose calculation models rely on raytracing and analytical kernels to represent dose deposition in matter. One of such models, the PB model, decomposes the continuous diverging radiation beam into a superposition of discrete beamlets.…”

SAR Image Simulations Using the LBM Algorithm on MPI-GPU

“…GPU task-parallel models have recently become a popular topic in GPU research. Aila and Laine (2009) presented the idea of persistent threads for handling irregular ray generation in ray tracing and this was further developed in a GPU ray tracer OptiX (Parker et al 2010).…”

Exploring heterogeneous computing with advanced path tracing algorithms

“…The 1536KB L2 cache is shared among all SMXs. The path tracing algorithms running on the NVIDIA devices run on top of the OpTiX ray tracing engine [16], [17]. The second used server is endowed with two such boards.…”

“…Each GPU accelerator present on the heterogeneous system is also assigned an iteration following the same demand driven policy. Besides the above explained memory limitations, the ray tracing engine used within the GPUs is OpTiX [17], inhibiting the simultaneous execution of more than one iteration per graphics board. At this upper level of workload decomposition and scheduling, tasks -iterations -are strongly related to memory address spaces: computational resources sharing the same address space (cores within the central processing unit, cores within the Xeon Phi, elementary processors within the GPU) are assigned a single iteration, such that memory consumption, i.e., the number of photon maps, does not increase with the number of such resources per device (we have in fact one photon map per device).…”