Why do VR games need so much more power to run?


One thing that’s throwing me for a loop is that VR games run at fairly normal resolutions. For instance, the oculus rift S only has a 1440p screen, it’s the same 2560x1440p resolution. A GTX 1070 or higher can run most games at that resolution at 90fps or above. So with worse lighting quality what makes these games so demanding to run with the same hardware?

In: Technology

They don’t actually require much power than a PC game running at the same resolution and settings. For example I have a gtx970 and it can run rift games without issue.

But VR companies typically recommend higher specs than needed for two reasons.

1) It’s a new tech. You don’t want people to think the tech is garbage because their PC is. If someone *very first time* using VR was a terrible experience they are a lot less likely to come back. That’s not the case with normal video games anymore.

2) any decrease in frame rate for even a brief second can cause someone huge amounts of disorientation and make a game unplayable. That’s not really the case on in regular games. A small drop in frame rate is annoying but not going to make you throw up, which it can in VR.

To achieve an authentic 3D effect, VR games need to render two separate images for every frame, one for each eye. So straight off the bat you need nearly double the horsepower for the same resolution on a single monitor.
Also there is a lot of extra positional tracking data which is being processed simultaneously.

Former game developer here,

The biggest problem is all the overhead in rendering the scene twice. You have to transform all the geometry of all the polygons in the scene, twice; you have to cull and occlude, twice; you have to UV map your textures, twice; you have to compute all your lighting, twice; you have to alpha, raster, dither, and anti-alias, twice.

Bigger screen geometry has nothing to do compared with having to do the same work twice. A bigger screen simply means more rasterizing, and that’s already one of the faster things a GPU can do. And in the end, you have a novel effect that really isn’t that much more compelling than rendering a 3D projection onto a 2D screen once. You can get an adequate immersive effect by having mono vision in an enclosed headset simply because it’s like putting blinders on a horse, and you’re the horse. If you were told it was VR, you might not even be able to tell the difference.

VR is a novelty that has been popping up since the 80s, again and again. They even tried to bake it into panel televisions in the early 2000’s, and that went over like a lead balloon. The industry is mostly not interested in the technology because the visual effects are twice as expensive but not for twice the gain, but, as you’ve noticed, you definitely lose color and lighting – that’s always been a problem inherent with the technology. The major studios would rather deliver you one spectacular render 60fps than two mediocre renders that don’t add to the gameplay.

So the technology will always remain niche. It’s a novelty. And what no one is willing to admit is there is a big liability issue because no one knows what VR does, long term, to the young, developing minds of children. Sega developed a whole VR console, it was to replace the Genesis (it might have been what turned into the 32x), and the testers noticed quite a few extra fender benders in the office parking lot. Legal pulled the product because they didn’t want to get sued for harming a child, and there is no way to ethically perform such a study. We do know it does have lasting effects on the brain of adults.