Khronos Releases ANARI 1.0 Provisional API for Scalable 3D Data Visualization
Beaverton, OR – November 2, 2021 – Today, The Khronos® Group announces the release of the ANARI™ 1.0 (Analytic Rendering Interface) provisional open standard API for scalable 3D data visualization. ANARI enables users to build the description of a scene to generate imagery, rather than specifying the details of the rendering process, providing simplified visualization application development and cross-vendor portability to diverse rendering engines, including those using state-of-the-art ray tracing. In addition to the ANARI specification, Khronos has released a sample implementation, starter applications, developer tools and conformance tests into open source – and welcomes community feedback before specification finalization in 2022.
3D visualization is a vital tool in supercomputing and across the industry to understand, illustrate, and gain insights from complex data. Before ANARI, visualization engines, libraries, and applications needed to be ported to proprietary APIs to access rendering engines from multiple vendors. Designed for high-performance within today’s supercomputing environments, ANARI will save domain and visualization experts significant development time by enabling them to tap into the power of advanced rendering techniques, including ray tracing and distributed renderers by simply building a description of a scene, rather than dealing with multiple low-level rendering engines. This provides visualization application portability to a wide range of backend rendering engines, which are in turn free to innovate their use of rendering technologies, including the latest techniques such as ray tracing.
The scientific visualization ecosystem includes key visualization application vendors such as VMD, VTK/ParaView, and VisIt that have actively participated in ANARI’s design. There are also early ANARI implementations in development by AMD, Intel and NVIDIA that will provide access to their Radeon ProRender, OSPRay, and VisRTX rendering engines respectively.
“A win-win for the industry, ANARI is designed to enable scalable, portable rendering that makes state-of-the-art rendering techniques and hardware-optimized renderers widely accessible, while still enabling the interactivity necessary for exploratory visualization,” said Jefferson Amstutz, ANARI working group chair and senior software engineer at NVIDIA. “Khronos anticipated this industry need and has been working on the ANARI specification for over two years to bring the 3D visualization community a well-designed, cross-platform API. Looking beyond our initial focus on scientific visualization, we believe ANARI will provide value to many developers across diverse domains that need a simpler, high-level API to render sophisticated imagery.”
ANARI is platform-independent, C99/C++-based, and designed to enable scalable rendering to use any available distributed compute and rendering resources. Direct coupling with applications enables asynchronous overlap of compute and rendering, and large memory-resident volume and geometry data sets can be processed in-situ. Runtime querying enables “matchmaking” of applications to suitable renderers.
ANARI 1.0 has been released as a provisional specification to enable the incorporation of industry feedback before the specification is finalized. Khronos welcomes comments and feedback from the data visualization community at the specification GitHub repository to ensure that ANARI integrates effectively with existing workflows.
In addition to the specification, the ANARI Working Group has released a collection of open source software under the Apache 2.0 license to help developers familiarize with the API and implementations to become officially conformant when the final specification is available in 2022. This software includes: a simple ANARI implementation with CPU-based ray tracing and no external engine dependencies, a library of small example ANARI applications, an early validation layer for debugging and tracing API calls, and a beta release of the ANARI conformance test suite.
Industry Support for ANARI:
“The ANARI rendering API is a big step in standardizing the pipeline of scientific rendering, allowing developers and scientists to write code once and render many places without needing to know the specifics of the device. We look forward to supporting multiple platforms for GPU and CPU rendering via our Radeon ProRender backend for ANARI, which is expected to be available soon at GPUOpen.com,” said Brian Savery, Radeon™ ProRender software lead at AMD.
“As an advocate for open, industry-wide standards, Intel embraces the high-level rendering API specification ANARI led by the Khronos Group. The ANARI working group found the fine balance between providing ease-of-use for applications development and delivering capabilities for high-performance implementations by vendors. Intel supports the ANARI API with an open-source implementation based on Intel® OSPRay (part of the Intel® oneAPI Rendering Toolkit), contributing expertise and technologies to advance scientific visualization, high-fidelity rendering, and MPI scalability to ANARI users,” says Jim Jeffers, senior principal engineer and senior director of Intel Advanced Rendering and Visualization.
“At Kitware, providing access to the latest large scale scientific and medical visualization techniques is critical to our business. With ANARI we can leverage its API to easily support multiple backends ranging from ray tracing to traditional polygonal engines across deployments ranging from individual clients to distributed rendering on the largest supercomputers. For applications like ParaView ANARI is a huge win, giving us access to many rendering engines without the prohibitive per-engine development costs,” says Ken Martin, co-founder and distinguished engineer at Kitware.
“ANARI not only will contribute to application portability of visual analytics tools, but also will reduce the burden of developing and maintaining visualization services across pre-exascale and exascale systems at Oak Ridge Leadership Computing Facility (OCLF),” says Benjamín Hernández, computer scientist, Oak Ridge National Laboratory.
“The ANARI API will be a huge productivity booster for us, simply by reducing our development and maintenance costs quite significantly and by providing instantaneous portability across multiple hardware platforms,” says Christiaan Gribble, director of high performance computing, applied technology operation, SURVICE Engineering
“The ANARI open rendering interface standard will help address the unique needs of technical and scientific visualization tools, enabling them to leverage state-of-the-art renderers using powerful ray tracing methods for interactive, in-situ, and post-hoc visualization on hardware platforms ranging from PCs to the largest parallel computers in the world. ANARI’s high-level interface will enable visualization software developers to rapidly deploy cutting-edge hardware-optimized renderers and rendering technologies in their tools, relying on an ecosystem of robust standard-conforming implementations. Users of leading visualization tools such as VMD will be able to make routine use of state-of-the-art visualization performance and cinematic quality rendering within their preferred software,” says John Stone, senior research programmer at University of Illinois.