This sponsored post from Intel explores how the Intel Rendering Framework can deliver better performance at a higher degree of fidelity — without having to invest in extra hardware.
The Intel Rendering Framework brings together a number of optimized, open source rendering libraries. (Photo: Shutterstock/By Chutima Chaochaiya)
GPU or CPU?
A few years ago, the only way to get real performance for graphics rendering was with specialized add-on GPU hardware. Today the focus has shifted back to the CPU––especially modern high-throughput cores packaged in multi- and many-core processors and running software that can take advantage of high thread counts, large local memory and data parallelism.
By bringing the work back to the CPU, users are seeing better performance than they could get from dedicated graphics hardware and limited memory. But the key ingredient is software optimized for Intel Xeon® processors, like the Intel® Rendering Framework.
The world of virtual design for the automotive, aeronautics, and manufacturing is just one example of industries that rely on high visual fidelity photorealism and interactive rendering.
Performance and scalability are not only relevant for scientific visualization, but also for cinematic and professional visualization applications. Here, with CPU memory typically on the order of 192 GB or more versus only 16GB or less in most GPUs, the Intel Rendering Framework can deliver better performance at a higher degree of fidelity. And, by letting the CPU do the work, visualization applications can run anywhere without specialized hardware.
To achieve this degree of high-performance, high-fidelity visualization, the Intel Rendering Framework brings together a number of optimized, open source rendering libraries:
- Intel® Embree, a ray tracing kernel library adopted by more than 80 independent software vendors for its extensibility, and vast feature set for complex rendering and shading frameworks.
- Intel® OSPRay, a cluster-capable, scalable, extendable, ray-tracing, rendering toolkit that includes path tracing and volume rendering. It efficiently uses threading and vectorization to create interactive, high-fidelity applications on Intel CPUs.
- Intel® OpenSWR, an OpenGL* low-level rasterization library upstreamed to the Mesa OpenGL open source project to achieve high rendering performance when GPUs are unavailable or are too limiting. Completely CPU-based and built on top of LLVM, this library provides performance advantages over Mesa llvmpipe renderers.
- Intel® Open Image Denoise, an open source library of image denoising and filtering algorithms to improve visual quality and reduce rendering times.
A key feature of this software-defined visualization approach is that it now makes in situ rendering possible. This puts rendering on the same nodes as the computation, keeping data in memory, avoiding storage transfers and setting a path for visual data analysis at the exascale level.
With no special hardware requirement, high-performance CPU-based rendering with the Intel Rendering Framework becomes available to nearly every researcher or video analyst on most available platforms, from laptops and workstations to enterprise clusters, world-class supercomputers and even in the cloud.
It’s been shown that as few as eight Intel Xeon Scalable 8180 processors can provide the computing power needed for quality interactive photorealistic ray tracing. Applications can easily scale up to big data and exabyte simulations, view results on their laptops, switch to display walls, or in fully immersive cave experiences. This ability to run and visualize on CPUs anywhere and without specialized hardware brings professional rendering applications into the HPC datacenter.
The world of virtual design for the automotive, aeronautics and manufacturing is just one example of industries that rely on high visual fidelity photorealism and interactive rendering. By adopting software-defined visualization technology, these application domains can achieve their quality and performance needs with scalable software modeling that uses both serial and parallel processing on the CPU.
The Intel Rendering Framework provides highly optimized visualization operations through known rendering APIs. Taking this further, applications now being developed converge the power of artificial intelligence and ray tracing to generate real-time animation characters that can learn body movements in photorealistic scenes.
The Intel Rendering Framework is one of a number of initiatives from Intel advancing the serial and parallel performance of many- and multi-core Intel Xeon processors to reduce the need for specialized hardware. Experiencing the spectacular images created using the Intel Rendering Framework shows what software-defined, CPU-based visualization can produce.
The Intel Rendering Framework open source libraries are available for download here.
