nVidia Interview - Why G-Sync Is Superior to FreeSync

[nimbulan]

I thought I'd post this here, an interview with nVidia's Tom Petersen about what makes G-Sync better than FreeSync and why he believes that it's worth the extra cost. Whether or not you agree with him it's a pretty interesting read about what goes into providing a good gaming experience with variable refresh rate monitors.

http://www.forbes.com/sites/jasonevange ... -freesync/

[GameLifter]

That's a great read! It's not too surprising that G-Sync has a leg up over Freesync at the moment but it comes at a cost. Even though I already own a G-Sync monitor I do hope Freesync/Adaptive Sync improves over time. The more VRR tech the better!

[sharknice]

Great article, thanks for linking that. I think a lot of people speculated some of the things he said but that definitely confirms things.

[lexlazootin]

Thats interesting, does anyone know if the GSYNC mode fixes the intense over shoot of Overdrive? because ATM that's one of my main complaints of my Benq monitor.

[nimbulan]

Thats interesting, does anyone know if the GSYNC mode fixes the intense over shoot of Overdrive? because ATM that's one of my main complaints of my Benq monitor.

From my understanding the G-Sync modules are tuned for each monitor to provide the best image quality whether G-Sync mode is enabled or not. I know on the Asus VG248QE original scaler, you could turn up the overdrive to the point of serious overshoot but the G-Sync module uses a setting low enough that that does not happen.

[RealNC]

Why does my bullshit-alert-spidey-sense tickle when I read stuff like:

Q: "What benefits are there for [insert scripted question here]."
A: "That's an interesting question. [insert scripted answer here while pretending that the question wasn't known beforehand]."

[Sparky]

Why does my bullshit-alert-spidey-sense tickle when I read stuff like:

Q: "What benefits are there for [insert scripted question here]."
A: "That's an interesting question. [insert scripted answer here while pretending that the question wasn't known beforehand]."

Well, I'm not going to deny it was a softball interview, but you can't expect a forbes guy to know enough about g-sync's limitations to properly grill Tom. I expect they both read the same Freesync reviews, and the questions and answers were pretty obvious. The only thing that made me stop and think is the idea that AMD needs to customize it's drivers to each panel. I don't see why that would be expected of AMD, if anything the necessary info could be put in the EDID, or other configuration data.

[sharknice]

Why does my bullshit-alert-spidey-sense tickle when I read stuff like:

Q: "What benefits are there for [insert scripted question here]."
A: "That's an interesting question. [insert scripted answer here while pretending that the question wasn't known beforehand]."

I don't think it was scripted, but he just let the nVidia guy say what he wanted. He never actually answered the how much does the GSYNC hardware cost question and instead spun it his way.

[jts888]

The interview does bring up the interesting point about the increased ghosting in the Adaptive-sync panels, but it does a very bad job discussing the underlying reasoning IMO.

I was under the impression that ghosting is mostly a product of back-end T-CON behavior, which has little to nothing to do with the scaler/DisplayPort-side frame input. It seems like the first round of Adaptive-sync/FreeSync monitors punted on using agressive panel overdrive, but there is nothing about Adaptive-sync that makes this inherently impossible, or even necessarily difficult.

G-sync is a full end-to-end platform including a monitor-side specific hardware implementation, while Adaptive-sync is just a standardized communications protocol with implementations being theoretically able to be identical, better, or worse than G-sync's, but this interview does not make that clear whatsoever.

[nimbulan]

The interview does bring up the interesting point about the increased ghosting in the Adaptive-sync panels, but it does a very bad job discussing the underlying reasoning IMO.

I was under the impression that ghosting is mostly a product of back-end T-CON behavior, which has little to nothing to do with the scaler/DisplayPort-side frame input. It seems like the first round of Adaptive-sync/FreeSync monitors punted on using agressive panel overdrive, but there is nothing about Adaptive-sync that makes this inherently impossible, or even necessarily difficult.

G-sync is a full end-to-end platform including a monitor-side specific hardware implementation, while Adaptive-sync is just a standardized communications protocol with implementations being theoretically able to be identical, better, or worse than G-sync's, but this interview does not make that clear whatsoever.

I believe the point he's trying to make is that the amount of overdrive needed is dependent on the refresh rate, hence the G-Sync module keeping previous frames in memory and adjusting overdrive calculations on a per-frame basis. Every G-Sync monitor will have this capability but most FreeSync monitors will not.