Intel significant release of scalable 3D rendering engine , designed for realistic high quality ray tracing rendering suitable for use in interactive applications. The engine is being developed as part of a larger project , aimed at developing software visualization tools for scientific calculations (Software Defined Visualization), including the ray tracing library , photorealistic rendering system , a library for denoising images and software rasterization system . The code is written in C++ and licensed under Apache 2.0.
OSPRay is primarily aimed at interactive applications for rendering scenes on the fly. The method used to simulate the behavior of light is . Visualization is supported in volume and on a plane, photorealistic taking into account the physical properties of materials, advanced shading effects (shadows, transparency and shading "").
OSPRay uses only the capabilities of the CPU, not being tied to the GPU, which allows the library to be used on a wide range of devices, from workstations to nodes in computing clusters. To ensure proper performance, multithreading and vectorization based on SIMD instructions, such as Intel SSE4, AVX, AVX2, and AVX-512, are actively used (OSPRay requires at least SSE4.1 support).
Rendering can be distributed over several nodes of the cluster (supported by MPI), which, for example, allows you to use OSPRay to organize rendering of a picture with a very high resolution on , a single image on which is formed by a set of separate LCD panels. For example, OSPRay is shown in a composite screen , composed of 80 30-inch monitors (total resolution 40960 Γ 8000 or 328 megapixels) and served by a cluster of 40 servers with 6-core CPUs based on the Intel Sandy Bridge microarchitecture.
A significant change in the version number is due to a large reworking of the API, including changes that violate compatibility (to simplify the transition to the new API, a layer library has been proposed that smoothes migration), and the provision of new geometric types. Added support for Open VKL (Open Volume Kernel Library) for volumetric rendering. Implemented the ability to connect to reduce image noise. The ospcommon library and the MPI support module have been moved to separate repositories.
Source: opennet.ru