Truth about Nvidia's G-sync Marketing Module.

[atari314]

NVIDIA using VESA Adaptive Sync Technology – Freesync.

 

http://gamenab.net/2015/01/26/truth-about-the-g-sync-marketing-module-nvidia-using-vesa-adaptive-sync-technology-freesync/

[AlwaysFSX]

Well there's really only one way to do the same thing..

[brownninja97]

WWWWWWHHHHHHHHHHHAAAAAAAAATTTTTTT

 

 

well honestly i dont care, as long as i benefit from it

[Real_PhillBert]

Repost:

http://linustechtips.com/main/topic/299222-is-nvidia-g-sync-a-scam/

 

Also I'm not buying it.

[Peter Vaughan Truslow]

NVIDIA using VESA Adaptive Sync Technology – Freesync.

 

http://gamenab.net/2015/01/26/truth-about-the-g-sync-marketing-module-nvidia-using-vesa-adaptive-sync-technology-freesync/

I'd love to see some third party confirmation of this working. I won't install a modded driver though until it has been confirmed.

[Sauron]

NOW WAIT A FREAKING MINUTE.

 

 

Which Monitors are compatible with the G-sync Modded Drivers ?

 

All the external monitors that include DP 1.2 (past 2010) and all the Laptops that include eDP.

 

are you telling me my asus pb287q, which does not officially support gsync or freesync, nor the 1.2a DP update, ACTUALLY supports it and all that's stopping me from using that potential are drivers? What? Wasn't there a monitor firmware impediment?

[Guest]

NOW WAIT A FREAKING MINUTE.

 

 

are you telling me my asus pb287q, which does not officially support gsync or freesync, nor the 1.2a DP update, ACTUALLY supports it and all that's stopping me from using that potential are drivers? What? Wasn't there a monitor firmware impediment?

 

I really would like to know this as well. If my new Dell 2414H (DP 1.2) supports Adaptive-Sync, then I want a FreeSync/G-Sync video card. I get that NVidia is using Adaptive-Sync technology, because there's probably not a ton of ways to achieve this, but I don't see why they want more royalties than AMD with their FreeSync.

[Sauron]

I really would like to know this as well. If my new Dell 2414H (DP 1.2) supports Adaptive-Sync, then I want a FreeSync/G-Sync video card. I get that NVidia is using Adaptive-Sync technology, because there's probably not a ton of ways to achieve this, but I don't see why they want more royalties than AMD with their FreeSync.

 

they want more royalties because they're nvidia jokes aside, the way I understood it is that nvidia actually supplies g-sync capable scalers, before run-of-the-mill scalers supported adaptive sync. So the "brain" of the monitor is actually made by nvidia, which would sort of justify the royalty (although after adaptyve sync became a thing they should have just adopted it, but that's beyond the point). This guy instead is saying that vesa stealth-included some sort of adaptive sync into the DP standard before gsync was presented and before they publicly announced they'd include the feature in the 1.2a revision. It sound pretty fishy to me. Besides, if the monitors are not the obstacle, I don't see why GPUs would be. Why shouldn't, say, a 280x support freesync if adaptive sync was included in DP since 2010?

 

I wouldn't get my hopes up to be honest, there are too many things that don't make a ton of sense. IF IT'S TRUE HOWEVER, that day we'll drink a beer at the same time staring at each other's avatars.

[Guest]

they want more royalties because they're nvidia jokes aside, the way I understood it is that nvidia actually supplies g-sync capable scalers, before run-of-the-mill scalers supported adaptive sync. So the "brain" of the monitor is actually made by nvidia, which would sort of justify the royalty (although after adaptyve sync became a thing they should have just adopted it, but that's beyond the point). This guy instead is saying that vesa stealth-included some sort of adaptive sync into the DP standard before gsync was presented and before they publicly announced they'd include the feature in the 1.2a revision. It sound pretty fishy to me. Besides, if the monitors are not the obstacle, I don't see why GPUs would be. Why shouldn't, say, a 280x support freesync if adaptive sync was included in DP since 2010?

 

I wouldn't get my hopes up to be honest, there are too many things that don't make a ton of sense. IF IT'S TRUE HOWEVER, that day we'll drink a beer at the same time staring at each other's avatars.

 

If the video card supports DP 1.2, the monitor as well, and Adaptive-Sync is part of DP 1.2, then it *should* be only a question of driver. To be honest, I'm not excited about G-Sync/Freesync, but if the feature is available for my hardware, then why not enabling it (if it works fine).

 

Looking at your avatar for too long might cause epileptic seizures. 

[Sauron]

If the video card supports DP 1.2, the monitor as well, and Adaptive-Sync is part of DP 1.2, then it *should* be only a question of driver.

 

but that IS the problem right? dp 1.2 should NOT support adaptive sync  :huh:

[Guest Generallee]

but that IS the problem right? dp 1.2 should NOT support adaptive sync  :huh:

 

That's what they wanted you to believe

 

WAKE UP SHEEPLE !!1!

[Sauron]

That's what they wanted you to believe

 

WAKE UP SHEEPLE !!1!

 

^^ I don't see a logical reason for VESA to do that though

[Glenwing]

The article is wild speculation and wrong on many levels lol. But to humor them I downloaded the modified driver since I have a G-SYNC capable graphics card (780 Ti) and three different DisplayPort monitors (two of which are revision DP 1.2a), and as expected it did not work, no G-SYNC option with a Dell U2414H, U2415, or ASUS PA248Q over DisplayPort. None worked. Yes I completely uninstalled old drivers first, yes I made sure the monitors were in DP 1.2 mode, yes I disabled all OC profiles and ran at stock GPU settings, and I tested the displays one at a time and didn't use daisy chaining or SLI. I wanted to give this a fair shot at working, with no weirdness in my setup. It did not work on any of the monitors, although I already knew it wouldn't.

 

Ultimately G-SYNC and Adaptive-Sync ARE hardware solutions, don't let anyone fool you into thinking otherwise. The display's refresh is initiated by the display controller, the GPU has no power to control it directly. On a display that doesn't have dynamic refresh capability it's actually impossible to get G-SYNC working with just a driver hack. If the display controller is not programmed for a variable refresh frequency, there's no amount of driver or software magic on the source side that can make it so. The very least possible requirement would be a firmware update on the monitor side. But most monitors wouldn't handle it very well anyway, in the end you really need a display controller designed to refresh dynamically. On a laptop with eDP's dynamic refresh power-saving feature implemented... maybe. Big maybe. But on any desktop monitor? Not a chance.

 

(I hope no one minds I'm going to promote this to best answer myself for the sake of making sure people aren't misinformed...)

[atari314]

Actually I do mind.

 

Also: "Update #6 : Recorded video at 240fps without any modification by youtube

-snip-

Edited January 28, 2015 by Blade of Grass

[D2ultima]

NOW WAIT A FREAKING MINUTE.

 

 

are you telling me my asus pb287q, which does not officially support gsync or freesync, nor the 1.2a DP update, ACTUALLY supports it and all that's stopping me from using that potential are drivers? What? Wasn't there a monitor firmware impediment?

I can confirm the Clevo P570WM with two 980Ms in it can use Gsync due to DP 1.2... I can also confirm that if your external display is wired to your intel GPU (due to Optimus) you will not be able to use it. Good luck finding out the truth! =D.

[atari314]

"Update #7 : Official Geforce drivers vs G-sync Modded Drivers (holding my smartphone to show in realtime that G-sync work)

https://www.youtube.com/watch?v=qZYfnpH3JWg"

 

^ is this a misleading link too? How much is nvidia paying for PR dmg control?

[RichardF]

Very interesting.

 

Something like this should be part of the standard monitor spec in the first place. We shouldn't have to rely on proprietary stuff to get decent 3D quality.

[RichardF]

I just read his latest comment. It sounds like his argument now is that there needs to be a database in the driver that has settings that are compatible with each monitor model. He says the differences between monitors is the reason why his modified driver won't work for most people.

 

Like I said, the main problem is that, my algorithm is kinda old and need to be improve to support a large amount of monitor. It was based on eDP monitor, thinking that It would be universal and will work with DP 1.2 monitor, I was wrong, the main problem is that. The controller inside each monitor can be completely different, driving same panel specs is not the only matter, you also need to respect the schematic of the controller itself, it is the most important part of a panel, it is the one who decide yes or not what can be executed.

 

Again I was wrong thinking that I would work with the algorithm I made in this condition, timing controller is why adaptive-sync, G-sync, Freesync etc.. exist. Why a small amount of monitor are G-sync and Freesync and not more of them ? The monitor is not the problem, every monitor as the capacity to hold this.

 

Let’s take for example NVIDIA and G-sync, G-sync is exactly adaptive-sync but the difference is that, NVIDIA make it compatible with the brand they deal with, the algorithm itself will be adapted to a specific amount of monitor that NVIDIA choose, the G-sync algorithm include all the necessary call function that deal with the controller available on the monitor, this has nothing to with G-sync module, the G-sync module is a programmable component that valid the calling function, if the G-sync module is not here, you will face to a return error. When I was talking about DRM, I was talking about how NVIDIA take care of the process in here instead of releasing a fully free software way for the people who has a NVIDIA GPU.

 

AMD is on the same boat, again Freesync = Adaptive-sync, AMD do exactly the same thing, saying is free, well is not totally free since, since you need to pay for it, so you can use their algorithm that support the amount of monitor they choose to deal with. Now do you understand ? the main problem is not the monitor itself, it is to deal with a large amount of different timing controller, Kind of preset or if you want let’s say like calibrating your monitor by using some ICC profile.

 

Let’s take fast another example, on the Geforce drivers and AMD drivers, you have inf file that include all the needed information about your GPU card and how to deal with each of GPU (unified drivers), each GPU card has a different call function, they don’t work the same way, so the drivers is here for that, saying to the API where to find the information that the API want, each GPU need to respect API standard for example DX or OGL but also need to come with the drivers that will deal with API (working together hand to hand with the GPU) drivers are here to control the GPU (included a lot lot of different algorithm). Here is the same for the monitor, Geforce drivers include G-sync algorithm and how the GPU can work with monitor, that’s why we need the display driver to deal with external command/data (GPU as the master) respecting each interface and controlling each pulse to show the frame on the monitor. You have a way to offer for free that type of technology because in here is not the hardware part the problem is the software part.

 

Unfortunately I’m not a team and I can’t not work alone on everything, I do my best, take my time to be helpful... I will continue to work on it and will come back later, EDID software mod can be helpful, remade the algorithm to support more and more monitor controller (even more helpful) and try to open this thing out, so people can exploit their monitor and GPU as it was supposed to be. Again I still think this technology should be free to share (people already buy expensive GPU and monitor) and now they need to deal with a problem that has been made by company who made GPU and monitor, think about it.

There is more but I think this covers the main idea.

[Badger906]

I have it.. don't think it's worth the premium.. probably getting an amd card next.. so won't be using it

[Majestic]

I think websites are using these tumultuous time to fire more clickbait on the web. They seek to aslo benefit from Nvidia's mistake. Don't go to these websites, don't click their fucking clickbait articles. 

 

Really, learn to think for yourselves.

[atari314]

oh look, guess what's real: https://www.youtube.com/watch?v=K7SYvgB6SZ4

now, coff coff, where are all the ppl saying that was fake?

 

also: "This seems like some shady BS to me.. Like the module is just there as a form of DRM. It seems like you guys spent most of the video doing damage control pr for Nvidia, or you're trying to avoid getting sued by them. This works exactly as normal G-Sync works, Why are you acting like proper desktop G-Sync looks good at 25fps? Then you explain the importance of the module being invented? lol. It looks like Nvidia got caught out on a bit of a lie here.. Maybe give some credit to the guy that found this out?" --Gorilla Reckoning

 

[ItsAFeature]

Please stop posting the conspiracy theories.

 

1. Laptops like the ones used here have eDP which is capable of dynamic refresh. Regular monitor scalers are not. Hence the new DP 1.2a scalers that are just now coming to market. There was no hardware in monitors to support G-Sync at the time it first came out. Nvidia made their own scaler to do just what they needed for G-Sync.

 

2. The guy who "found this out" got a leaked driver from the ASUS forums. He then tried to cover up the fact that it was an accidentally leaked prototype driver of G-Sync for mobile devices. PCPer took issue with him claiming to have hacked the Nvidia drivers when all he did was install a driver directly from Nvidia.

 

3. People weren't saying that eDP G-Sync wasn't possible. We were saying that eDP G-Sync does not translate to G-Sync for any monitor with a DP connection.

[RichardF]

Please stop posting the conspiracy theories.

Like that Nvidia sold a card with incorrect specs, OCZ switched the NAND from 32-bit to 64-bit without telling anyone in the Vertex 2 (leading to a big loss of performance and also a loss of capacity), and a leading ssd maker recently was found to have swapped out the controller it used in the drives it sent to review sites with the cheaper slower Sandforce controller?

 

[Glenwing]

oh look, guess what's real:

now, coff coff, where are all the ppl saying that was fake?

 

also: "This seems like some shady BS to me.. Like the module is just there as a form of DRM. It seems like you guys spent most of the video doing damage control pr for Nvidia, or you're trying to avoid getting sued by them. This works exactly as normal G-Sync works, Why are you acting like proper desktop G-Sync looks good at 25fps? Then you explain the importance of the module being invented? lol. It looks like Nvidia got caught out on a bit of a lie here.. Maybe give some credit to the guy that found this out?" --Gorilla Reckoning

 

The majority of it was fake, as it turns out. Don't try to draw attention away from all the other crap the guy said. The best reply to this was already posted by the creators of the video you just linked in their full review.

 

January 30, 2015 | 03:56 PM - Posted by Allyn Malventano

We had no issue with, and even named Gamenab as the first to discover this. What I *do* hold issue with, however, is that he tried to pass it off as his invention, making all sorts of claims that the module uses crypto, etc. It was too easy to see straight through his act.

• His first video attempt showed a game running *windowed*. Anyone with a shred of adaptive sync experience knows that it can only properly work full screen.
• Simple analysis of his 'modified' driver revealed a .sys file lifted from another driver build.
• When people started calling him on his sillyness (myself included), he started taking offense and only dug himself deeper with conspiracy theories about how *his* algorithm magically got into an Nvidia driver.
  • ...and that specific driver just happened to be leaked by a tech support rep
  • ...posted about on a forum thread
  • ...relating to the laptop that he just happened to already own.

Seriously, that guy needs to buy a lottery ticket if he was that lucky.

 

http://www.pcper.com/reviews/Graphics-Cards/Mobile-G-Sync-Confirmed-and-Tested-Leaked-Alpha-Driver/Digging-Deeper

 

With a new driver G-SYNC does work on some specific laptops, that much turned out true, but the guy made claims that went far beyond that, most of them patently absurd and just plain wrong including his basic premise (that the G-SYNC module's only purpose is to act as DRM confirmation). It's based on THOSE that I (and most others) decided he was full of shit (which he is), not because we thought it was impossible to implement G-SYNC without a module.  I'll quote myself again for consistency:

 

On a laptop with eDP's dynamic refresh power-saving feature implemented... maybe. Big maybe. But on any desktop monitor? Not a chance.

 

So as I predicted, it's possible for just a driver to enable variable refresh on a laptop, but not on desktop monitors. And how did things turn out? What do you know, the driver works on a few laptops, but not on desktop monitors like the guy claimed it did (after claiming to have modded it himself, when he didn't). So yes, the original article was full of bs from some guy looking for attention for work he didn't do.

In any case, I am locking this thread as we don't need more than one. Continue any discussion here: http://linustechtips.com/main/topic/301268-mobile-gsyncits-gonna-be-a-thing/