Preamble: Technical Lingo
A quick word upfront: this blog post contains some technical terms and abbreviations, which might not be well-known to every reader. To assist the reading and
understanding, [here at the bottom of this post] is a short glossary of some technical terms used in this article.
Max FPS. Max Quality. Powered By AI.
The demand for stunning visuals in gaming and serious digital experiences continues to rise, and developers have always faced the challenge of balancing image quality with performance. That is… until now. Breakthrough technologies like DLSS (Deep Learning Super Sampling) and DLAA (Deep Learning Anti-Aliasing) have come into play as true game changers. Both powered by advanced AI (Artificial Intelligence), these solutions deliver impressive improvements in rendering quality and efficiency, for the first time ever breaking the barrier of having to trade off one for the benefit of the other. By harnessing deep learning, DLSS upscales lower-resolution images to better-than-native quality while also boosting frame rates. DLAA focuses on maximizing image quality with an AI-based anti-aliasing technique. Together, they represent a new era of rendering technology that offers the best of both worlds: stunning visuals with high performance, making them game changers for both digital entertainment and industrial applications.
DLSS uses deep learning algorithms to upscale lower-resolution images into higher-resolution ones while maintaining high visual fidelity. It essentially renders fewer pixels and then uses AI to create the rest, allowing for smoother frame rates without loss in image quality.
Key benefits:
- Performance boost: DLSS helps applications run at higher frame rates by rendering fewer pixels and using AI to upscale the images.
- Image quality: DLSS (especially DLSS 2.0 and beyond) produces native image quality or even better.
- Real-time AI processing: NVIDIA Tensor Cores (present in NVIDIA RTX GPUs) power DLSS, allowing AI models to perform complex upscaling and anti-aliasing in real time.
Application For Industrial Use Cases
Beyond gaming and entertainment, technologies like DLSS and DLAA are proving transformative in industrial applications, particularly in VR (Virtual Reality). For industries like automotive, aerospace, high tech, product design and many more, the demand for photorealistic virtual prototypes has never been higher. The key challenge is achieving the high frame rates required for a smooth VR experience while maintaining visual fidelity. DLSS addresses this by boosting frame rates through AI-powered upscaling, making it possible to implement real-time ray tracing without compromising performance. NVIDIA RTX ray tracing, known for its ability to render physically accurate reflections on surfaces, becomes more feasible in VR environments, unlocking new use cases such as visibility studies, material testing, and interactive design reviews. This combination of technologies allows engineers and designers to interact with true-to-life digital prototypes, enhancing decision-making and reducing the need for costly physical mock-ups.
Harnessing Cutting-Edge Technology
As a key innovator in our field, we at Dassault Systèmes follow a long tradition of cultivating a technology collaboration with NVIDIA, to develop and integrate hardware and software and to maximize the benefits of digitalization for our users. Our 3DEXPERIENCE platform utilizes several of NVIDIA’s advanced technologies, ranging from AI-powered image denoising for global illumination rendering to the support of the latest NVIDIA RTX GPUs for maximum performance and even real-time ray tracing.
The applications stretch far beyond simply enhancing the performance and visual quality for our users: by combining the most advanced hardware and software, we even unlock new use cases for Design & Engineering, which previously seemed out of reach. With NVIDIA’s DLAA, DLSS and RTX real-time ray tracing, combined with the natively integrated push-button VR on the 3DEXPERIENCE platform, our users can experience physically correct reflections on any surface as part of their visibility checks and interactive ergonomic assessments.
We continuously evolve our rendering in collaboration with NVIDIA. To learn more about NVIDIA RTX and the latest supported drivers on the 3DEXPERIENCE platform, visit:
Dassault Systèmes certified drivers on NVIDIA website
Continue your journey and discover our dedicated CATIA portfolio all around high-end rendering, interactive product experiences and visual decision-making for designers and engineers here.
And join the CATIA Design & Styling community, free login
Glossary: The Technical Lingo Explained
Aliasing happens when you use pixels to draw lines or curves on a screen (also known as rasterized graphics). It’s the dreaded “stair steps effect” which causes undesired image artifacts and poor image quality. Aliasing also happens when tiny or distant objects are smaller than one pixel on the screen, causing the outlines of objects to flicker, which can be especially distracting in VR.
Anti-Aliasing (AA) is the name for the different methods to alleviate aliasing artifacts, by using interpolated color gradients for the surrounding pixels. While some AA methods process the image in native resolution to smooth out sharp edges, others require the source image to render at increased resolution, which is then scaled down (this especially helps to improve sub-pixel size artifacts). The challenge for all AA methods is to smooth out artifacts, without blurring or smudging the overall image. AA often requires a tradeoff between the desired quality and the available performance.
Deep Learning (DL) is a subset of machine learning and describes the training of artificial neural networks to process data. “Deep” refers to multiple, stacked layers of artificial neurons in the network. All the currently popular AI models for natural language assistants or image generation are based on Deep Learning.
Frames Per Second (FPS) describe the number of images (frames) rendered and displayed per second, which determines the overall performance and smoothness of a visual experience. The importance of high FPS grows linearly with the overall interactivity of your scene; maximum FPS are required for an enjoyable virtual reality experience with stereoscopic 3D rendering. Another critical factor for performance, especially for industrial applications, is the typical use of large data sets (large assemblies and native, precise CAD geometry), which increases the demand for memory and computation capacity. While computer games typically employ heavy optimizations like limiting the polygon count and using pre-calculated LODs (Level of Detail) for their assets, industrial applications require working on native, live data sets in fast-paced environments, where designs constantly change without much time for content optimization.
Graphic Processing Unit (GPU) is specialized hardware architecture initially designed and optimized for the fastest rendering and display of image information. Some computers have integrated (“onboard”) GPUs, but the most powerful ones are usually separate units (“graphic cards”) with their own cooling and dedicated memory and core architecture. Thanks to their extreme real-time processing power, today’s GPUs are also used for other live processing of heavy data, like analysis of sensor inputs or training models for artificial intelligence.
NVIDIA RTX™ Ada Generation is the latest and highly decorated GPU ecosystem by NVIDIA. Offering the most advanced platform for ray tracing and AI technologies, RTX revolutionizes ways to play and create. Leading games and applications use RTX to deliver realistic graphics, incredibly fast performance, and new cutting-edge AI features like NVIDIA DLSS. Our 3DEXPERIENCE platform supports NVIDIA RTX GPUs natively.
Super Sampling (SS) means that the source image is rendered at a higher resolution than the desired target output. The high resolution image offers higher fidelity and more detail, which is then scaled down to interpolate the details for a smooth result. It is a powerful method for Anti-Aliasing, but it also requires lots of computation power.
Virtual Reality (VR) is considered the pinnacle of real-time visualization. It brings a stereoscopic, “true-3D” experience, which requires the application to render twice the load of FPS (different perspectives for left and right eye of the user). Stereoscopic VR is usually combined with tracking the user in real time with six DOF (Degrees of Freedom) for capturing rotation and translation in 3D space, to accurately translate all the user’s movements into the virtual scene. This creates a virtual 3D environment for the user to explore 3D CAD models in a fully spatial, life-like experience. VR offers the maximum advantage in Design & Engineering to understand sizes and proportions, visibility and ergonomics and spatial relations without physical mock-ups. High FPS are essential for a convincing and enjoyable VR experience. The 3DEXPERIENCE platform offers native, push-button VR in all of our 3D applications, supporting most Virtual Reality devices available on the market via the OpenXR standard.