VESA Adopts Adaptive-Sync

[Mand]

The CES demo did show it working. Watch it again. The left laptop has v-sync enabled. The FPS changes slightly, but the frame time value ie the refresh rate does not move except for 1/100th increments about 3 times. The right shows the FPS and frame time changing rapidly by up to 1/10 increments. Neither is fixed, but both are syncing. It's not 50hz v-sync or you'd see the same as the left where the frame time is static only changing in the 1/100 value range. The only issue with the demo is the right laptop's render isn't complex enough to show large swings in FPS to make it as obvious as Nvidia's demo.

The fluctuations you describe are what you get whenever you turn vsync on, regardless of what framerate you set it to. Fluctuations of a tenth of a FPS are not indicative of variable refresh, they are indicative of imprecision in the framerate counter. Framerate counters count the time interval between frames, and then invert it for frames per time.

I can get fluctuations well above 0.1 FPS with vsync on on my current, aged machine, that certainly doesn't have variable refresh.

Rapid fluctuations slightly below 50 is what 50 Hz vsync looks like. It's what I get when I stare at a blank wall in a game with vsync on, fluctuations in the 59.7-59.9 range, occasionally reading 60. Framerate counters are not precise enough for these fluctuations to be significant, and they're certainly not indicative of a variable refresh rate.

There's a reason the render isn't complex enough to show large swings in framerate. Because the demonstration isn't showing variable refresh. It shows two STATIC refresh rates, with vsync on. Normal, standard, decades-old vsync.

 

[Phynaz]

Some info. Note dynamic per frame refresh rate.

Summary
Extend the "MSA TIMING PARAMETER IGNORE" option to DisplayPort to enable source based control of the frame rate similar to embedded DisplayPort.

Intellectual property rights
N / A

Benefits as a result of changes
This enables the ability for external DisplayPort to take advantage of the option to ignore MSA timing parameter and have the sink slave to source timing to realize per frame dynamic refresh rate.

Assessment of the impact
The proposed change enable per frame dynamic refresh rate for single stream devices that expose dynamic refresh rate capability in EDID for DisplayPort interface. The source will be able to enable this with an SST interface or MST hub with physical ports. Logical MST port support of the feature is not included as part of this SCR. A generic framework to enable such feature for logical port is required that can accommodate other feature where stream related configuration is programmed in DPCD.

Analysis of the device software involvement
SST device which support "MSA TIMING PARAMETER IGNORE" option will be able to expose the capability in EDID and DPCD to let source enable dynamic refresh rate.
Source driver would have to be updated to parse EDID and enable "MSA TIMING PARAMETER IGNORE" feature when source want the sink to be refreshed based on its update rate.

Analysis of the compliance test and interop implications
Currently this feature is tested as part of eDP CTS. New test would have to be added as part of DP LL CTS and EDID CTS.

 

[Mand]

Yes, I know that's what they say. What I'm saying, and what you seem to be having trouble accepting, is that they haven't shown it.

I can say I built a teleportation system in my spare bedroom, but nobody will care unless I can demonstrate it.

There is way, way, way more involved than just having a timing parameter that has the capability to carry a variable refresh signal. The meat of the technology is actually doing the variable refresh, and that's what this spec doesn't cover and AMD is hoping other people figure out.

Just because you include something in the signal to allow for variable refresh does not mean that variable refresh just starts happening. It's one necessary component of the solution, but it is not itself the solution.

 

[monstercameron]

Yes, I know that's what they say. What I'm saying, and what you seem to be having trouble accepting, is that they haven't shown it.

I can say I built a teleportation system in my spare bedroom, but nobody will care unless I can demonstrate it.

There is way, way, way more involved than just having a timing parameter that has the capability to carry a variable refresh signal. The meat of the technology is actually doing the variable refresh, and that's what this spec doesn't cover and AMD is hoping other people figure out.

Just because you include something in the signal to allow for variable refresh does not mean that variable refresh just starts happening. It's one necessary component of the solution, but it is not itself the solution.

I don't get any of your posts...are you saying that adaptive-sync is a lie or that it doesn't work? Are you saying VESA are idiots who would add a null feature to the spec?

 

[Phynaz]

So you're back to saying the VESA membership is either lying or incompetent.

You're never going to be convinced otherwise, so I'll step out now.

Have a great weekend.

 

[Mand]

I'm saying that AMD itself said the point of pushing for a standard update was to encourage hardware manufacturers to develop the technology. Direct quote.

There's no guarantee that it will actually pan out, given that. There is a very, very big difference between "We have this technology, we'd like you to put it in the DisplayPort spec so we can sell it" and "We're putting this in the DisplayPort spec and hope that people develop the technology." A big, big difference. Especially since they've released no information whatsoever, no evidence whatsoever, on something they make a big, big, BIG deal about how open it all is.

It is most certainly not enough for you to proclaim "G-Sync is dead" in the OP.

I would be THRILLED to be convinced otherwise, and I've said, repeatedly, what the numerous ways I could be convinced otherwise entail. Any one of the following would satisfy me:

1) A working demo
2) A technical description of the technique involved and requirements for hardware
3) Announcement of specific hardware partners

So far we have nothing but an optional VESA spec that if the entire planet proceeded to ignore, would have absolutely zero impact whatsoever on anything we could get on our desks. Standards die on the vine all the time. Remember HD-DVD? It was the competitor to Blu-Ray, and it failed. It was also the "open, industry standard" version compared to the proprietary Blu-Ray, owned by Sony. The proprietary standard won that fight. And HD-DVD was even far enough along that you could buy discs and players for it (and can now get really, really cheap on ebay). If G-Sync wins, if no hardware manufacturers pick up A-Sync and do the requisite hardware development to implement it, then A-Sync will be nothing more than a line in a Word document. Just because it's in a VESA spec does not mean it's absolutely going to happen. There are lots of features that get approved, but then don't become popular. It's entirely possible A-Sync can fall into that category.

If you think this means I think they're lying or incompetent, then that's your business. I don't, but if that's what it takes for you to dismiss what I say, I suppose I can't stop you.

 

[SoulWager]

I don't get any of your posts...are you saying that adaptive-sync is a lie or that it doesn't work? Are you saying VESA are idiots who would add a null feature to the spec?

As I understand it, AMD misrepresented the CES demo, and no working implementation of refresh rate adapting to a changing framerate has been demonstrated to the public, except for g-sync.

Don't be confused by PR speak. Wait until hardware gets sent to reviewers for thorough testing and direct comparison to g-sync.

 

[Leadbox]

snip

You know what the tech is intended for, you've been given a timeline (6 to12 months).
You want a running commentary on their current progress?

 

[Mand]

A running commentary would be fantastic, but any indication at all of any progress whatsoever, even once, would be sufficient.

They've said they're working with hardware partners, but haven't said who they are.

If they're working with hardware partners, then that means that they probably have some idea of the technical approach, yet won't say what it is.

If they have some idea of the technical approach, but won't say what it is, why the big hype from them about open standards and no royalties or licensing? Why hide what they have done so far? Why haven't they had any specific information to present?

Put all this together - no demo, no technical info, no announced partnerships, no hardware manufacturers saying they're going to make a product with it - and I question whether they have anything at all. And not unreasonably, I don't think.

I'd love to be proven wrong. Maybe in 6-12 months they'll pull a rabbit out of their hat, but before I can get behind statements like "G-sync is Dead" I'm going to need at least some evidence that this isn't vaporware. Anything at all would do, but so far there's just nothing.

 

[stahlhart]

This discussion has run its course. Thread closed.
-- stahlhart