NVIDIA Demonstrates Experimental “Zero Latency” Display Running at 1,700Hz

[ahhming]

 

 At GTC 2016, Nvidia demoed a prototype display that runs at about 1700Hz.

Nvidia attached a VR display to a rail system and shook it back and forth rapidly. The image on the display stays in the same position relative to the camera. The logo stays visible with only a small amount of blurring, despite the rapid movement. The presenter, Nvidia’s Vice President of Research David Luebke, says that Nvidia attains this high refresh rate by "overloading the display and driving it with this kind of novel binary delta-sigma encoder kind of approach." The company didn't indicate when this display tech might make its way out of the lab.

 

Source:

http://www.roadtovr.com/nvidia-demonstrates-experimental-zero-latency-display-running-at-17000hz/

[AHaskin14]

That's quite fast

[SurvivorNVL]

nVidia G-Sync + nVidia display technology = nVidia display port 7.0, can you run 4K at 2000hz?  

In all seriousness this is extremely interesting, but I doubt zero-latency.  There's likely some latency, no matter it's probably extremely low, and 1700hz is incredible if true.  I'd like to see them get that on an actual monitor.  That'd be a great exclusive reason to keep buying nVidia.  G-Sync, OLED, 1ms Display-latency across the board, 200-1700hz refresh rate.

[DocSwag]

Zero latency? Cool, though I'd like to see some real testing for that.

1700 hz? Too be honest, I don't give a crap about that. Why? Because it's pretty obvious that we can't see 1700 hz. The old thing about us only being able to see 24 fps is definitely pretty untrue, but I think that in VR our goal should be in the 200-300 fps range. That would look pretty fluid to me. 1700 is just a total waste.

[Ren]

This will be the best technology available to you if you're ever in a prolonged earthquake. Great work NVidia.

[Majestic]

1 hour ago, DocSwag said:

Because it's pretty obvious that we can't see 1700 hz. 

Yes, that is the bloody point

[FruitBasketSilex]

1 hour ago, DocSwag said:

Zero latency? Cool, though I'd like to see some real testing for that.

1700 hz? Too be honest, I don't give a crap about that. Why? Because it's pretty obvious that we can't see 1700 hz. The old thing about us only being able to see 24 fps is definitely pretty untrue, but I think that in VR our goal should be in the 200-300 fps range. That would look pretty fluid to me. 1700 is just a total waste.

It's likely that they were just expirementing, found a way to hit higher refresh rates, and then found just how high their new method could go with 1700hz. I doubt they were explicitly meaning to get 1700hz. This could however possibly bring down prices of higher refresh rate displays once they actually refine it into a product.

 

Also, we need a GPU capable of 1700fps.

[raphidy]

Oh god. NVIDIA headset on the way.

[NateGSR117]

That really cool! That a really fast refresh rate that I certainly won't be able to tell the difference. But I sure this tech will down scaled down to more reasonable 144hz to 244hz on an IPS (or even OLED) panel, please?

[dalekphalm]

35 minutes ago, Majestic said:

Yes, that is the bloody point

No, his point was that, we can't see 1000 Hz either. Or 750 Hz. Or probably 500 Hz.

 

I would say that from a purely technical level, there's likely little need to ever go above 500 Hz (Hell, even 250 Hz is probably sufficient for decades to come).

 

I understand that they likely showed off 1700 Hz because that's the magic number that their R&D tech capped out at, not that we needed 1700 Hz specifically. The zero latency portion is much more interesting, but the 1700 Hz portion is basically useless. Give us true 240 Hz, instead of a useless 1700 Hz.

[SurvivorNVL]

9 minutes ago, dalekphalm said:

No, his point was that, we can't see 1000 Hz either. Or 750 Hz. Or probably 500 Hz.

 

I would say that from a purely technical level, there's likely little need to ever go above 500 Hz (Hell, even 250 Hz is probably sufficient for decades to come).

 

I understand that they likely showed off 1700 Hz because that's the magic number that their R&D tech capped out at, not that we needed 1700 Hz specifically. The zero latency portion is much more interesting, but the 1700 Hz portion is basically useless. Give us true 240 Hz, instead of a useless 1700 Hz.

I believe there was a test that was used, using a display tech primarily for fighter-pilots in the U.S., and the screening has various levels of increasing refresh-rates, but it's a low as all hell resolution; one in thirty people can notice 250hz+, one in fifty 400hz+, and one in a hundred can notice 600hz+.  So, yeah, given Linus did an old NCIX video of measuring to see if he can tell the difference between 120 and 144, and he knew every time, where as a normal person couldn't tell the difference just being seated down by it - I can imagine this is largely useless, unless it can help solve motion sickness, due to some micro-issues.  Our eyes pick up strobing that we physically can't see on PWM displays, so maybe such a high refresh rate is needed for VR.  

[2FA]

Zero latency is physically impossible for a display, just saying.

[DXMember]

oooor.... you could just run CRT?

[DXMember]

8 minutes ago, DeadEyePsycho said:

Zero latency is physically impossible for a display, just saying.

if you want to talk what's physically impossible then "zero latency" itself is physically impossible regardless of what sentence the phrase is used in

[2FA]

51 minutes ago, DXMember said:

if you want to talk what's physically impossible then "zero latency" itself is physically impossible regardless of what sentence the phrase is used in

The reason I put displays is because the study of quantum mechanics has already changed much of what we understand in physics, for all we know there could be some form  of zero latency. When it comes to our current knowledge of electricity though, you are correct that there is no such thing as zero latency.

[DXMember]

1 minute ago, DeadEyePsycho said:

The reason I put displays is because the study of quantum mechanics has already changed much of what we understand in physics, for all we know there could be some form  of zero latency. When it comes to our current knowledge of electricity though, you are correct that there is no such thing as zero latency.

what do you think brain works on? pixie dust?

[2FA]

1 minute ago, DXMember said:

what do you think brain works on? pixie dust?

That literally has nothing to with what I was saying. Have you even read about any of the observations physicists have made while studying quantum mechanics? Stop diverting the topic.

[ThinkWithPortals]

4 hours ago, ahhming said:

binary delta-sigma encoder kind of approach

I love how sci-fi that sounds

[Sharkyx1]

2 hours ago, FruitBasketSilex said:

Also, we need a GPU capable of 1700fps.

You don't need your GPU to be as fast as your screen really, and this would effectively be a Gsync Display without any fancy daughter boards or software, as its intervals 1-144fps would likely be smaller than a single frame

[Maslofski]

3 hours ago, DocSwag said:

Because it's pretty obvious that we can't see 1700 hz. 

 

wxactly, but ive they can get it running ok on 1700hz then a 300hz display should be a super reasonable task for that technology

[Bensemus]

Wouldn't you also need a gpu capable of outputting 1700Hz? To me it seems the max a gpu can output is it's clock speed. Or maybe double it's clock speed if it sends an image on the rising and falling edge. If the second is true then we are good xD.

[DXMember]

22 minutes ago, DeadEyePsycho said:

That literally has nothing to with what I was saying. Have you even read about any of the observations physicists have made while studying quantum mechanics? Stop diverting the topic.

physics is about observing the physical world, we are the observers, we have inherently wired latency in our mechanism of observing.

it's not about what's possible for quantum particles to achieve, it's about what's possible for us to observe and interact with

 

P.S. topic is about displays

[Stefan1024]

I can controll an RGB LED with a PWM of >1 MHz.

Give me a good FPGA and I will make you a display ith 1 000 000 FPS and ~10x10 pixels.

[That Norwegian Guy]

2 hours ago, dalekphalm said:

No, his point was that, we can't see 1000 Hz either. Or 750 Hz. Or probably 500 Hz.

Actually, the fact that humans can see lightning strikes proves that we can perceive the equivalent of at least 800hz.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[2FA]

34 minutes ago, DXMember said:

physics is about observing the physical world, we are the observers, we have inherently wired latency in our mechanism of observing.

it's not about what's possible for quantum particles to achieve, it's about what's possible for us to observe and interact with

 

P.S. topic is about displays

And what does our own built in latency have to do with the latency of a display?