AMD demos FreeSync (free G-Sync alternative) [AnandTech]

[TheExodu5]

http://www.anandtech.com/show/7641/amd- ... t-ces-2014

Using two Toshiba Satellite Click notebooks purchased at retail, without any hardware modifications, AMD demonstrated variable refresh rate technology. According to AMD, there’s been a push to bring variable refresh rate display panels to mobile for a while now in hopes of reducing power consumption (refreshing a display before new content is available wastes power, sort of the same reason we have panel self refresh displays). There’s apparently already a VESA standard for controlling VBLANK intervals. The GPU’s display engine needs to support it, as do the panel and display hardware itself. If all of the components support this spec however, then you can get what appears to be the equivalent of G-Sync without any extra hardware.

Good to see some competition. But this begs the question: why did NVidia go with a far more expensive hardware approach if a standard was already in place that could accommodate variable refresh rates? I'm thinking there has to be a downside to this.

[nimbulan]

AMD's demo makes it sound like you can just plug any old monitor in with a firmware update and it will work like G-sync does. The fact that they did the demo on a laptop display which I believe relegate more control to the GPU coupled with the fact that nVidia's having to make $200 upgrade kits that replace most of the monitor's electronics to accomplish this tells me that it's really not going to be any different. I was very surprised to read that AMD hardware has supported this feature for 3 generations, yet nVidia produced a product first.

I suppose we'll have to wait for more information, and confirmation about compatible standalone monitors to know for sure.

Edit: The upgrade kits are $200 now that nVidia has revealed the price.

[sharknice]

I have doubts this will work on any existing stand alone monitors. I don't think VBLANK intervals were meant to be changed on the fly. I suspect when you change the interval the monitor would behave the same way as when you change other things like the resolution. It would black out for a second and readjust.

Just my guess, I haven't messed around with any of that stuff before.

[Chief Blur Buster]

I am especially interested how it scales to modern LCD's such as 1080p 120Hz displays.

[RealNC]

I'm interested at how quick the displays respond. NVidia are using a dedicated controller for this. Maybe they had good reason. If input lag with FreeSync is way worse than G-Sync, then G-Sync wins. If not, then NVidia is running a scam, selling us a solution to an already solved problem

[TheExodu5]

Here is a good read on the subject:
http://techreport.com/news/25867/amd-co ... -sync-tech

Koduri explained that this particular laptop's display happened to support a feature that AMD has had in its graphics chips "for three generations": dynamic refresh rates. AMD built this capability into its GPUs primarily for power-saving reasons, since unnecessary vertical refresh cycles burn power to little benefit. There's even a proposed VESA specification for dynamic refresh, and the feature has been adopted by some panel makers, though not on a consistent or widespread basis.

According to Koduri, the lack of adoption is simply due to a lack of momentum or demand for the feature, which was originally pitched as a power-saving measure. Adding support in a monitor should be essentially "free" and perhaps possible via a firmware update. The only challenge is that each display must know how long its panel can sustain the proper color intensity before it begins to fade. The vblank interval can't be extended beyond this limit without affecting color fidelity.

And finally:

In Koduri's assessment, it's possible to achieve a G-Sync-like animation smoothness with a combination of two techniques: dynamic refresh rates and triple buffering. He thinks Nvidia's G-Sync hardware is doing both of these things in order to achieve the results it does, but he initially expressed puzzlement over why Nvidia chose to implement them in expensive, external hardware. After all, triple-buffering can be implemented by a game developer in software or even enabled via a software switch in a graphics driver control panel. Koduri said AMD used to have an option to force the use of triple buffering in its driver control panel, in fact, and would be willing to consider bringing it back.

So it seems that NVidia may be using hardware for 1 of 2 reasons. Either their display controller cannot support dynamic refresh rates, or they wanted to avoid a solution that required triple buffering. Since AMD is implying that this solution requires triple buffering, this immediately brings one huge drawback into the picture: input lag.

It seems like FreeSync may provide a decent free and open alternative to G-Sync, so I'm still hoping NVidia ends up supporting it as well. Just treat G-Sync as the premium option that also eliminates input lag.

[nimbulan]

So it seems that NVidia may be using hardware for 1 of 2 reasons. Either their display controller cannot support dynamic refresh rates, or they wanted to avoid a solution that required triple buffering. Since AMD is implying that this solution requires triple buffering, this immediately brings one huge drawback into the picture: input lag.

It seems like FreeSync may provide a decent free and open alternative to G-Sync, so I'm still hoping NVidia ends up supporting it as well. Just treat G-Sync as the premium option that also eliminates input lag.

I'm curious why they are claiming triple buffering is necessary. The whole purpose of triple buffering is to allow the GPU to continue rendering while waiting for the monitor to refresh which will not be necessary when the monitor always refreshes immediately after a frame has been drawn.

[TheExodu5]

It could be that the video card needs to measure the time difference between two frames to control the length of the vblank interval. I'm under the impression that using this method, the video card needs to predict ahead of time how long the monitor will have to hold a certain image, so it needs to look ahead by at least 1 frame.

[Chief Blur Buster]

I'm curious why they are claiming triple buffering is necessary. The whole purpose of triple buffering is to allow the GPU to continue rendering while waiting for the monitor to refresh which will not be necessary when the monitor always refreshes immediately after a frame has been drawn.

It still takes a finite amount of time for a display to refresh. Scanout time for Freesync is currently 1/60sec, so that blocks that buffer. Triple buffering eliminates the blocking overhead, and reduces input lag, which is important.

One conceptual example of triple buffering:
(1) Front buffer that the display is reading from, while display is scanning out the refresh.
(2) Back buffer waiting to flip to refresh.
(3) Rendering buffer, which replaces whatever is in the back buffer.

At high framerates, (3) will run many times, keeping (2) very fresh (low input lag), so that when the refresh is finished (2) will be flipped to (1) for the next refresh pass at the lowest possible latency.

[Chief Blur Buster]

It could be that the video card needs to measure the time difference between two frames to control the length of the vblank interval.

This wouldn't be necessarily if you just simply kept continuously transmitting black front porch scanlines. A properly implemented VRR technology (Variable Refresh Rate) does not need to know in advance, when the next refresh is going to occur.

However, it's wholly possible that this is a necessary part of existing LCD chipsets, and this might be what is being done.