Quantum computing could boost Ray Tracing Performance By Up To 190%

Submit

Researchers across the United States, Portugal, and the United Kingdom predicted that the solution to the hefty performance requirements of ray-tracing might be a combination of older ray-tracing algorithms mixed with quantum computing. In a recently published research whitepaper, quantum computing enhanced ray tracing workloads, increasing performance by up to 190%. This process is done by limiting the number of computations needed by each ray.

Quantum Computing to increase complexity in ray-tracing technology

Ray-tracing in graphics technology has allowed for an evolutionary leap in gaming, especially with how the game titles are rendered. However, the performance and the developers' ability to adopt the process correctly have been minor compared to the complexity. The problem lies in the hardware and computational requirements of the ray tracing technology and the necessity for specific hardware that limits most users from the core technology.

Intel AXG Custom Compute team begins shipments of Blockscale ASIC, but is it too little too late?

Recently, AMD FSR 2.0, NVIDIA DLSS, and Intel's next-gen XeSS upscalers neutralize the higher performance disadvantages that follow the use of activating ray tracing in hardware. The individual upscalers minimize the number of pixels rendered to limit the formulaic intricacies of a specific scene before rebuilding the image to the required output resolution.

The researchers describe how quantum computing could potentially minimize the processing taxes caused by ray tracing technologies. The group took a 128 by 128 image processed with ray-tracing enabled and optimized the image using three different strategies. The three processes were classical rendering techniques, quantum rendering that is not optimized, and then optimization of quantum rendering. The first technique computed 2,678 million ray intersections on the 3D image, offering 64 per individual ray. The nonoptimized approach cut the first number by half, only needing 33.6 ray intersections, equalling 1,366 million ray intersections. Utilizing the optimized quantum technique joined with the classical system, the last attempt rendered the image with 896 thousand intersections with 22.1 rays each.

The most significant downfall to the technique was the quantum computing system. Quantum computers and devices are currently under development in the NISQ, or Noisy Intermediate-Scale Quantum product category. These intricate systems are not the highest in performance, so the rendering demands several hours to compute each image correctly. This category is perfect for simulations, but it is hardly a viable option at this time for rendering games.

Even though the results were excellent, the technology is far from available for production. With the current trend of quantum computing in the last year to two years, we are only seeing a small amount of quantum computing available for use. IBM plans to increase the volume of quantum computing in the upcoming years, but it is unknown how far the technology will advance in a short amount of time.

Linux Gamers Should Stick With Mesa’s RADV Drivers For RDNA 2 GPUs As It Outperforms AMD’s Official AMDVLK Vulkan Driver

Time and cost are not allowing for the technology to be markedly wide for the consumer market space. However, with the advancements made to cloud gaming in the last few years alone, the end-user may see this technology come sooner rather than later.

Source: Towards Quantum Ray Tracing: A Preprint (PDF), available through the arXiv at Cornell University

Submit