Virtual Reality Finds A Superhero In The Closet

A faster and more efficient VR technology will come into the market courtesy the use of GPUs for rendering codes and data.

David Kosslyn and Ian Thompson, the founders of Angle Technologies, lead a stealth project that may well render the current tech supporting Virtual Reality devices, on their silicon heads.

Backed by USD 8 million in funding, the duo have been keeping a low profile and an even tighter lid on their progress. Unwilling to disclose any details that would indicate the magnitude or the capabilities of the VR technology that they’re building, they did share some insights that they’d distilled from their trial and error efforts to increase the efficiency and the quality of their VR projections.

According to them, the method they uncovered has already been in use in other industries, but when used in the VR industry, it will alter the relationship between computer hardware and software as we know it. They go as far as hypothesising that it might be the future of VR programming!

What these two developers are using is a frequent life-saver in Game programming. And here is how it can be the life-saver for VR programmers too:

• The Graphics Processing Unit (GPU) is what game developers have been using since the 2000s for rendering their pixel-intensive games. Similarly, important software like Photoshop and Corel Draw also require a strong GPU for better performance.
• At Angle, while GPUs primarily tend to providing better graphics to the users, it is also being used for a number of other important tasks.
  Kosslyn and Thompson are reportedly shifting several other tasks onto these chips because GPUs are very efficient when it comes to running a horde of calculations in parallel.
• Because these chips can execute tasks and codes simultaneously, juggling through hundreds of digital tasks at the same time, both the developers believe the GPUs have the potential to steer VR faster than usual, to the intended goal of high quality and efficient visual services — even to the extent that in search of an illusion, it has the capability of transporting you to an ‘alternate dimension’.

When weighed against the CPU’s abilities to render all these elements, GPUs do even more.

Kosslyn and Thompson recollect how they were primarily using the CPU to install a plethora of elements individually like trees, bushes, leaves etc, when they were hit with the inevitable — loading each element took as much as one fifth of a millisecond too much. Considering the fact that close to a million such elements have to be rendered while developing and while in use by the user, the time taken to just load the world, would have been extremely slow.

For the user, to wait, for a program for so long, would be a development catastrophe. Hence, the usage of the GPU became a necessity.

“We’re mostly trying to fill in those few milliseconds when GPUs can do work besides rendering graphics”, Thompson says.

Along with Angle, Nvidia, which is probably the most respected GPU producer, suggests that GPUs would be the best choice for VR. Even insinuating that the idea of using GPUs in VR headsets would be a boon, given hundreds of these chips can fit into it. And that might just be the answer as more graphics intensive programs are added to VR.

On the contrary, Intel which is Nvidia’s competitor, disagrees, primarily because it does not produce these chips, but has subsidiary companies which it bought working in the VR space.

But as we see it now, it is highly likely that GPU will be the answer to all future VR technology and that VR’s emergence in smartphones might be facilitated by the power of these unsung heroes.

Originally published at Chip-Monks.