Nvidia's New Server Thinks in 3-D So Your Phone Doesn't Have to

Example of the types of images the server produces.
Example of the types of images the server produces.

Nvidia (s nvda) is moving further into the business of selling hardware, rather than just chips, with its new reality server that uses its graphics processors to offer photo-realistic 3-D images. The server uses an array of GPUs and Nvidia’s iray software to serve up the realistic imagery associated with special effects in movies through the browser on your PC or phone.

The product is built on Nvidia’s 2007 acquisition of Mental Images, which made software that ran on x86 processors, and created stunningly realistic pictures for advertisers and films. By replacing the x86 chips with GPUs, Nvidia says it has sped up the process of creating a moving photo-realistic image from what used to take hours into seconds. And because it’s delivered through a browser, everyone from e-commerce sites to social networks can offer this level of imaging to its constituents simply by hosting a Reality Server in its data center.

The technology can add far more depth to avatars on social networks as well as create accurate representations of products, such as clothing, that can be viewed from all angles. Because the imagery is delivered via the web as a series of jpegs, it isn’t dependent on Flash or on the viewing device having  a powerful graphics processor. So a site can deliver 3-D imagery easily on mobile phones over a 3G or even a 2.5G connection. This is yet another way we can extend the functionality of a mobile phone to make it more PC-like.

Nvidia has benefited as computing gets more visual, despite manufacturing issues that have fried some of its chips, and as we try to move beyond a text-based web into one where images, graphics and touch help us navigate information, products such as the reality server, and GPUs themselves, are going to be more important.