Nvidia’s secret weapon to win in the race for the metaverse


While the metaverse is now a priority for many technology giants, Nvidia took advantage of its recent keynote dedicated to graphics technologies to flex its muscles against the competition. There’s reason: Nvidia has indeed embarked on the race for the metaverse by presenting its latest technologies to make this digital universe a reality. The company – like the entire sector – however faces a weighty problem: how to transform a 2D world into 3D without going through complex, demanding and – above all – slow processes.

The American giant may well have found the parade to improve this thorny operation. How? ‘Or’ What ? Thanks to a new approach to reverse rendering – the process of reconstructing 3D scenes from a handful of 2D images. With its new method, Nvidia is harnessing artificial intelligence to approximate the behavior of light as it is in the real world. Better still: with the approach developed by the teams of the technological giant, the whole process now takes place almost instantaneously.

According to Nvidia, “this technology could be used to teach robots and self-driving cars to understand the size and shape of real-world objects. […] or in the fields of architecture and entertainment to quickly generate digital representations of real-world environments that creators could modify and expand.” Something to interest metaverse players, who are currently working hard to establish themselves on this market of the future.

Revolutionary technology

The latest technology designed by Nvidia could well be the breakthrough the metaverse needed to give itself the means to achieve its ambitions. Until now, using traditional methods to create a 3D scene could take several hours depending on the complexity and resolution of the visualization. However, artificial intelligence has greatly shortened the process by exploiting a popular new technology called Neural Radiance Fields (NeRF).

These NeRFs are based on neural networks to represent and render realistic 3D scenes from a collection of 2D images. A NeRF effectively trains a small neural network to fill in the blanks by predicting the color of light radiating in any direction from any point in 3D space. Early NeRF models have already delivered quality renders in just minutes, but training them takes hours.

This is precisely where Nvidia’s research teams come in. They have developed an instant NeRF, which combines fast neural network training and fast rendering. According to Nvidia management, this is the fastest NeRF technique to date, with speedups of over 1,000 times in some cases. To develop Instant NeRF, Nvidia relied in particular on a new method of input encoding called multi-resolution hash grid encoding. This method is optimized to work efficiently on Nvidia GPUs.

“Instant NeRF could be as important to 3D as digital cameras and JPEG compression have been to 2D photography, dramatically increasing the speed, ease and reach of 3D capture and sharing,” says David Luebke, vice president of graphics research at Nvidia. It remains to be seen what Nvidia intends to do with this new technology which could well allow it to strengthen its weight in the still nascent market of the metaverse.





Source link -97