SIGGRAPH the most important international event focused on computer graphics and interactive techniques launches July 28 in Los Angeles. From the partnership between ACCA software and AMD, here’s a glimpse at Edificius “AIrBIM”, the first Machine Learning rendering tool totally integrated in an Architectural BIM software.

Immersive technologies are changing the way we design, play, learn and communicate. These aspects, together with other ground-breaking technologies, are the main topics showcased at the prestigious SIGGRAPH (Special Interest Group on graphics and interactive techniques) international event. The computer graphics conference, widely recognized as an interdisciplinary educational experience inspiring technological innovations throughout the entire computer graphics industry, is also considered as the most important international meeting for presenting the results of research work in the computer graphics fields with dozens of projects showcased and presented to over 15000 CG leaders, trendsetters, decision makers and innovators together with the latest hardware, software and creative services.

This year’s 46th edition will be hosted at the Los Angeles Convention Center, from July 28th to August 1st 2019: a five-day dive into the latest technology trends in CG, animation, VR and conferences regarding AR, VR, MR and on new technologies that can be used in exciting and powerful ways.

siggraph-2019

When talking about VR and 3D, the past few years have seen a considerable progress in viewing technologies that assist and accompany us in everyday life: not only through the development of existing ideas and tools, but also through extension in other areas of use too.

Then there’s Real Time Live!, the event that celebrates the best interactive news of the year, the best of real-time graphics, all presented in an exciting live competition.

AMD & ACCA software at SIGGRAPH 2019

On the 50th anniversary of its foundation AMD, the US company that in the last few years has been revolutionizing the world of CPU and GPU high-performance graphics and processing technologies, will be presenting a wide range of innovating technologies at booth 1441, in particular with enhanced developments that aim to make animations, video games and creative graphics in general even more visually engaging.

AMD’s Radeon™ Prorender, the PBR (Physically Based Rendering) rendering engine with RayTracing technology combined with Machine Learning denoising capabilities, now extends its use to the AEC sector too.

For architects, designers and their clients, who need photo-realism to understand lighting solutions better and see how the different materials will influence their design concept, adding advanced time-saving “ray tracing” tools to an Architectural Design software such as Edifcius becomes crucial for the AEC industry professional to keep productivity high without necessarily having to be a graphics artist guru.

There are a multitude of advantages. In fact, designing a building being able to produce photo-realistic simulations with the right lighting solutions and precise materials greatly improves the design workflow.

This is the context in which AMD’s partnership with ACCA software becomes reality with new tools in the way of producing photorealistic renderings for architecture. A technology that allows designers to create renderings without having to be an expert and without having to spend hours trying different settings to get the right result and above all without having to use any external image post-processing applications.

AMD will be showcasing Edificius and showing the results of an intense collaboration with ACCA software focusing on the new AIrBIM (Artificial Intelligence render BIM) feature, the first Architectural BIM Design software integrated with AMD’s Radeon™ ProRender Ray Tracing technology.

AIrBIM, rendering with Artificial Intelligence

The Edificius AIrBIM tool, thanks to AMD’s Radeon™ ProRender, uses an Image denoiser based on Artificial Intelligence (Machine Learning Denoiser) to “remove” noise from a partially rendered image. A method that drastically reduces processing time

Radeon-ProRender-AIrBIM

The Machine Learning Denoiser is capable of comparing noisy images with their completely processed counterparts and reconstruct the final render much faster than before and automatically.

AMD has “trained” its denoiser to compare noisy images with their fully processed counterparts, and to quickly and automatically reconstruct the fully rendered image. With the Machine Learning Denoiser, there is now no need to render the entire multitude of light rays involved in a given scene (rays) having to process hundreds or thousands of render passes: the render is now completed much faster because only fewer passes are necessary considering that the machine learning algorithms can “remove” the residual noise normally present in a partically rendered image.

In practical terms this means no need to worry about complex settings to be applied each time to improve the rendering quality while keeping an eye on the processing time; each scene is automatically optimized with Artificial Intelligence denoising capabilities. By simply pressing the “Render” button, the user can make the scene lighting adjustments, or even the scene view, “on the fly” and while the rendering process is still working.

AIrBIM… incredibly fast and easy to use

  • faster, automatic and intelligent renders, thanks to the innovative Machine Learning Denoiser
  • better performance in terms of reduced rendering times, especially when used in combination with the automatic load balancing between GPU/CPU and memory sharing features (only if used with AMD graphics cards based on VEGA GPU-architecture)
  • improved quality and photo-realistic renderings thanks to the use of PBR materials (Physically Based Rendering) workflow
  • extraordinary lighting and reflection effects, with advanced global lighting control and 360° HDR panorama photos for scene backgrounds with IBL (Image-based Lighting)

Edificius Architectural Rendering generated with AMD radeon ProRender

  • Edificius Architectural Rendering generated with AMD radeon ProRender
  • Edificius Architectural Rendering generated with AMD radeon ProRender
  • Edificius Architectural Rendering generated with AMD radeon ProRender
  • Edificius Architectural Rendering generated with AMD radeon ProRender
  • Edificius Architectural Rendering generated with AMD radeon ProRender
  • Edificius Architectural Rendering generated with AMD radeon ProRender
  • Edificius Architectural Rendering generated with AMD radeon ProRender
  • Edificius Architectural Rendering generated with AMD radeon ProRender