What's going on with DisplayPort 2.0?

[Hexxagone]

It's already finished being made and standardized in 2020 and yet there is no talk about this from any major tech channels, no monitor has it and Nvidia skipped it altogether on the 4000 series.

 

What's going on here? 

[Eigenvektor]

It always takes time for new standards to be adopted. You need to keep in mind that designing a graphics card takes multiple years, so the DP 2.0 spec most likely wasn't finalized by the time the design of the RTX 40-series cards was.

 

In addition to this, the additional bandwidth of DP 2.0 isn't really needed for most people. DP 1.4 has enough bandwidth to drive the resolutions and refresh rates most people are gaming at.

[Zando_]

From the spec, seems there's just no need for 2.0 for any current monitor. Why would companies either raise prices or eat into their margins in order to put in a connector there's no current use for? 

[-rascal-]

12 minutes ago, Hexxagone said:

It's already finished being made and standardized in 2020 and yet there is no talk about this from any major tech channels, no monitor has it and Nvidia skipped it altogether on the 4000 series.

 

What's going on here? 

 

Due to factors...that I don't know... it has been delayed for a couple years.

It was planned for 2020, then it got delayed until 2021...and then again until 2022.

 

Like DDR5, PCI-E 5.0, USB 4.0, it takes time before it becomes available....even on server or HEDT platforms.

And like +1 or +1.5 year before it becomes available for consumer market.

 

AMD has already certified some of their laptop RX 6000-series GPUs for DisplayPort 2.0 compatibility.

The upcoming RX 7000-series is supposed to be DisplayPort 2.0 ready as well.

 

It is just in the recent months, there are reports of devices gong through DisplayPort 2.0 certification, AND passing.

[johnt]

What do we need it for?

[AnonymousGuy]

2 minutes ago, johnt said:

What do we need it for?

4K 144Hz without DSC.  DSC isn't lossless and does create visual artifacts for me with text.

[IPD]

I really think it's just Nvidia taking the piss.  It's a known quantity of need, and there's 0 reason why it should have been included on the 40 series.  I call "bs" on the "development lead time".  Plenty of interface options become available before the display tech will catch up to it.  Like @-rascal-said, PCIE 5.0 interface is offered on many things (including the Titan laptop LTT just built)--even though the SSD's to take full advantage of it don't yet exist.

 

DP 2.0 being offered first by Team Red simply gives me yet another reason to defect to Team Red; just like how their use of AGP encouraged defections 20 years ago.

[Stahlmann]

18 hours ago, AnonymousGuy said:

4K 144Hz without DSC.  DSC isn't lossless and does create visual artifacts for me with text.

While it's technically true that it's lossy, that doesn't mean it's bad. In my experience there is no visual difference to a non-compressed connection. I've used a few monitors using DSC and i've never had a problem. What you're describing sounds like chroma subsampling.

DSC is like FLAC audio. It's compressed to save data, but comparabe to uncompressed in terms of the user experience.

 

Overall it's simply not needed yet. DP 1.4 and HDMI 2.1 have plenty of bandwidth to feed all the monitors you could dream of. I'm personally still hoping DP goes away alltogether like it is with TVs. There is simply no reason to use DP over HDMI if you have the choice.

[IPD]

I'd be ok with DP going away entirely.  But HDMI is gonna have to get the bandwidth up to at least 3x 3840x2160 @ 180hz display support--before that will be a realistic thing.

[Doobeedoo]

Well RX 7000 will have it, soon to be revealed. And really it's a big improvement. 4K 240Hz 10bit HDR no DSC nonsense. Or 1080p 1000Hz. We may even see some sick 8K monitors. Either way at least we will see 4K at minimum I hope for new flagships, because some large 1440p and some ultra large 4K monitors lately look horrible with such low PPI though.

I want Samsung make 4K 240Hz QD-OLED model with it.

[Glenwing]

DP 1.2 was published in 2010. We didn't see widespread implementation in monitors until around 2013. DP 1.3 was published in 2014, we didn't see HBR3 in graphics cards until the end of 2016 as I recall. HDMI 2.1 was published in 2017, and didn't see wide adoption until around 2021.

 

DP 2.0 was only published in 2019. 3 years is a pretty typical time for technology to be developed that implements the standard to reach consumers. So 2022 would be the *expected* time under normal circumstances, and given COVID hit right after the standard was published, extra delay beyond normal was also to be anticipated. To be honest I'm surprised AMD was able to include it in their next generation. I expected to be waiting 1 more year.

[AnonymousGuy]

5 hours ago, Stahlmann said:

While it's technically true that it's lossy, that doesn't mean it's bad. In my experience there is no visual difference to a non-compressed connection. I've used a few monitors using DSC and i've never had a problem. What you're describing sounds like chroma subsampling.

DSC is like FLAC audio. It's compressed to save data, but comparabe to uncompressed in terms of the user experience.

 

No, it happens in RGB / 4:4:4.  DSC is described as "visually lossless"  but that is a fancy way of saying "you hopefully won't notice".    And it's true that I don't notice most of the time, but under certain conditions I see artifacting in my tab bar.  I can drop down to 120Hz where DSC turns off and it instantly fixes it.

 

FLAC is truly lossless compression.  MP3 would be more comparable to DSC because it's "audibly lossless" at higher bitrates.

[Glenwing]

8 hours ago, AnonymousGuy said:

No, it happens in RGB / 4:4:4.  DSC is described as "visually lossless"  but that is a fancy way of saying "you hopefully won't notice".    And it's true that I don't notice most of the time, but under certain conditions I see artifacting in my tab bar.  I can drop down to 120Hz where DSC turns off and it instantly fixes it.

 

FLAC is truly lossless compression.  MP3 would be more comparable to DSC because it's "audibly lossless" at higher bitrates.

Would you mind posting a screenshot next time you think you see artifacts from DSC? I'd like to test it myself. (Would need to be in uncompressed image format, like BMP).

[Hexxagone]

DSC would be an interesting topic for an LTT video

[AnonymousGuy]

On 9/22/2022 at 9:19 PM, Glenwing said:

Would you mind posting a screenshot next time you think you see artifacts from DSC? I'd like to test it myself. (Would need to be in uncompressed image format, like BMP).

It doesn't show up in a screenshot, only on the monitor itself.  Hard to take a picture because of all the image processing that happens with a photo but there's a thin horizontal white line between the T and S in "RTX 40 S" .

 

 

It comes and goes depending on what is on screen.  If I drop to 120Hz it's fully gone.

 

[Glenwing]

3 hours ago, AnonymousGuy said:

It doesn't show up in a screenshot, only on the monitor itself.  Hard to take a picture because of all the image processing that happens with a photo but there's a thin horizontal white line between the T and S in "RTX 40 S" .

 

It comes and goes depending on what is on screen.  If I drop to 120Hz it's fully gone.

 

I mean I wanted the original screenshot (uncompressed) so I can run the image through a DSC encoder and examine the results.

 

Encoded pixels in DSC only have a dependency on a few other adjacent pixels in the same line and previous line, and it is not possible for them to be affected by pixels further down in the image, so if the artifacts are at the top and are changing based on the contents further down, it isn't a DSC encoding issue; perhaps some kind of additional processing specific to that monitor or manufacturer (which they may choose to use only when DSC is active). Hard to say.

[AnonymousGuy]

12 minutes ago, Glenwing said:

I mean I wanted the original screenshot (uncompressed) so I can run the image through a DSC encoder and examine the results.

 

Encoded pixels in DSC only have a dependency on a few other adjacent pixels in the same line and previous line, and it is not possible for them to be affected by pixels further down in the image, so if the artifacts are at the top and are changing based on the contents further down, it isn't a DSC encoding issue; perhaps some kind of additional processing specific to that monitor or manufacturer (which they may choose to use only when DSC is active). Hard to say.

So do you need the entire screen or just a subset around the artifact?

[Glenwing]

13 hours ago, AnonymousGuy said:

So do you need the entire screen or just a subset around the artifact?

Well preferably the whole screen just to be sure, but technically only the first few lines would be needed. Also I'm not at home right now so I won't actually be able to look at it until next week. (Again, would need to be an uncompressed image like BMP, such as using the Print Screen button and pasting into MS Paint.)