VESA Announces Version 2.1 of the DisplayPort Standard

[Glenwing]

Nothing to get too excited about, this is more like a completion of some items that were left unresolved in version 2.0. There are no new transmission speeds or notable new features. The main topics are:

• Cable certifications
• USB4 integration improvements

Cable Certifications

Version 2.1 incorporates the new certifications tiers (DP40 and DP80) that were announced by VESA back in February of this year, and also restructures the lower certification levels a little bit.

 

The DP40 and DP80 certifications validate cables at the UHBR10 and UHBR20 speeds introduced in version 2.0. The names refer to the bit rates; UHBR10 = 10 Gbit/s per lane = 40 Gbit/s aggregate, hence DP40 certifies cables for 40 Gbit/s transmission, and likewise DP80 is for 80 Gbit/s (UHBR20 speed or 20 Gbit/s per lane).

Quote

DisplayPort 2.1 has also updated the DisplayPort cable specification to provide greater robustness and enhancements to full-size and Mini DisplayPort cable configurations that enable improved connectivity and longer cable lengths (beyond two meters for DP40 cables and beyond one meter for DP80 cables) without diminishing UHBR performance. VESA certified DP40 cables support up to the UHBR10 link rate (10 Gbps), with four lanes, providing a maximum throughput of 40 Gbps, while VESA certified DP80 cables support up to the UHBR20 link rate (20 Gbps), with four lanes, providing a maximum throughput of 80 Gbps.

 

The new standard also reorganizes the lower certification tiers, which have been a bit confusing over the years. Back in the DisplayPort 1.2 days there were no tiers; cables were either certified or they weren't, and any certified cable was good for the maximum allowed speed at the time (HBR2, 21.6 Gbit/s, i.e. 4K 60 Hz). However, this was no longer true once HBR3 speed (32.4 Gbit/s, 4K 120 Hz) was introduced in version 1.3, so VESA added the "DP8K" certification in 2018 (here). The original certification level (HBR2 and below) was then referred to as "standard DisplayPort certification".

 

So during this time, cables could have either a "standard DisplayPort certification" (good for up to HBR2) or "DP8K certification" (up to HBR3). Then version 2.0 was published, introducing three new speed tiers (UHBR10, 13.5, and 20). No new cable certifications were included (although this is common practice, I'll explain more below), but VESA did state that any DP8K certified cable should also be good for up to UHBR10 speed. In the words of Anandtech from the DP 2.0 release:

Quote

At just half the data rate of full-fat DisplayPort 2.0 (and Thunderbolt 3), UHBR 10 is resilient enough that it can operate over standard passive copper cabling, and cables should have little issue reaching 2-3 meters. VESA has actually been preparing for this for some time now, and UHBR 10 aligns with their previously-launched DisplayPort 8K cable certification program [DP8K]; 8K-certified cables will be able to meet the signal integrity requirements for UHBR 10.

 

Now, we arrive at today, with the DP40 and DP80 certifications, with DP40 being (also?) for UHBR10 speed and DP80 for UHBR20. Given the above context, one might expect that there is really only 1 new certification, DP80, and that the previous "DP8K" tier has simply been renamed to DP40 to align the names a bit better. Strangely, no; DP8K still exists as a separate tier, and VESA has quietly dropped the line about DP8K being suitable for UHBR10, so it's back to being an HBR3-level certification.

 

Confused yet? Perhaps a table will help. Here's the cable certification lineup and the transmission speed each one is suitable for, as of version 2.1:

Cable Certification	Transmission Speed	Bit Rate	Example Format (Uncompressed)
Standard VESA-certified	RBR	6.48 Gbit/s	1080p60 HDR
	HBR	10.8 Gbit/s	1080p120 HDR
	HBR2	21.6 Gbit/s	4K60 HDR
DP8K	HBR3	32.4 Gbit/s	4K120
DP40	UHBR10	40.0 Gbit/s	4K144 HDR
DP80	UHBR13.5	54.0 Gbit/s	8K60
	UHBR20	80.0 Gbit/s	8K75 HDR

 

I don't really mind the new additions of DP40 and DP80; if I were to nitpick, I'm not a fan of switching between per-lane bit rates and aggregate bit rates when deciding the names of cables versus transmission modes. I wish they'd pick one or the other. But moving beyond that, I'm still not a fan of the disjointed naming in the lower certification tiers. I would much rather they rename them to "DP20" and "DP30", or something like that. This would associate them with HBR2 and HBR3 speeds both by name and by bitrate (21.6 Gbit/s and 32.4 Gbit/s) while also harmonizing the names with the new DP40 and DP80 additions. But, we get what we get.

 

I should also comment on the fact that these new certification tiers weren't included when version 2.0 of the standard was published. It might seem odd to release the new standard which introduces UHBR10, UHBR13.5, and UHBR20 transmission speed, but not provide any certification for cables that can reach these speeds. This is, in fact, perfectly normal. Again I point to the DP8K certification published in 2018, for the HBR3 speed introduced in DP 1.3 (in 2014). We even had another standard revision (DP 1.4 in 2016) before the cable certification was ready. HDMI 2.1 was published in 2017, but the Ultra High Speed cable certification (while announced along with the specification) wasn't actually finished being written until the end of 2020 I think it was.

 

The simple answer is, it takes way longer to design and validate silicon than something like a cable, so when you finish writing the core design requirements, you publish them. That way, the people who are going to spend the next 3–4 years designing monitor controllers and GPUs can at least get started while you're still working on the other parts such as cable requirements, which don't take nearly as long to implement and can be published later no problem.

USB4

There are some technical improvements to integrate DisplayPort into USB better:

Quote

Achieving a robust, end-to-end user visual experience remains the utmost priority for VESA’s DisplayPort specification, whether across a native DisplayPort cable, via DisplayPort Alt Mode (DisplayPort over the USB Type-C connector), or tunneled through the USB4 link. As such, DisplayPort 2.1 has tightened its alignment with the USB Type-C specification as well as the USB4 PHY specification to facilitate a common PHY servicing both DisplayPort and USB4. In addition, DisplayPort 2.1 has added a new DisplayPort bandwidth management feature to enable DisplayPort tunneling to coexist with other I/O data traffic more efficiently over the USB4 link.

Final Thoughts

It's likely graphics cards and other devices will skip straight to "DisplayPort 2.1" branding. Version 2.0 was a sort of intermediary document which I anticipate will be "skipped over", similar to DisplayPort 1.3; you never really saw any "DP 1.3"-branded devices; for the most part we went straight from DP 1.2 to DP 1.4 (I think there was a brief period when graphics cards were advertised as "DP 1.3 certified/DP 1.4 ready" because of the timing of the releases, but that was it). This was because the version 1.3 standard was quickly revised with version 1.4 only 2 years later (before the hardware design cycle was finished), so by the time devices got to market, DP 1.4 was the current standard and its changes were also incorporated in products. Like I said, DP 2.1 is basically a "completed" version of the DP 2.0 standard with no core changes (no new features requiring new silicon designs), so I expect the same thing to happen. To be honest, I'm surprised it wasn't named DP 2.0a.

 

As always, I will remind people that there are no "DP 2.1 cables" or "DP 2.0 cables" or "DP 1.4 cables". There are (well, hopefully soon will be) DP80 cables which are certified for 80 Gbit/s, DP40 cables certified for 40 Gbit/s, and... well, I'm just going to continue calling them HBR3 and HBR2 certified cables. I really don't care for the "DP8K" and "Standard VESA-certified" designations. But please don't use version numbers. DisplayPort standard version numbers are for document tracking. They are not a cable classification system (or for that matter, a device classification system, but that's another topic ;D). Some versions introduce multiple new speeds (version 2.0 introduced three), while other versions don't introduce any new speeds at all. The classification of cables does not align with the document version numbers in any way. Please don't refer to cables using version numbers. Thanks

 

Source: VESA

 

(Edit note: previously I commented that the lowest cable certification tier was called "HBR"; actually I believe this is an internal name that will not be used for actual labeling, so I have edited these comments out.)

 

[ZetZet]

So is this the rumored standard AMD RDNA3 already supports before it was announced? 

[porina]

37 minutes ago, Glenwing said:

The classification of cables does not align with the document version numbers in any way. Please don't refer to cables using version numbers. Thanks

Maybe I should add it somewhere as a feature request but I don't seem to be able to edit tables to add more cells in this forum and I don't have alternate software on my laptop to do it externally. I think it would be useful to see what resolutions are supported at what frequencies, with and without DSC.

 

DP and HDMI both have a similar problem. We know what port we have, but it isn't obvious what cable is needed for a given display. I suppose it is a safe choice to pick a cable that is at least as good as the display. In a similar way, when I was shopping for certified HDMI cables to go with my then new TV, I knew I needed a cable for HDMI 2.1. The correct name Ultra High Speed HDMI cable might as well not exist. A potential problem with using the standard number for the cable also is future standards may not advance requirements but version mismatch would be confusing. It seems like every naming decision in tech has problems one way or another!

[Eigenvektor]

44 minutes ago, ZetZet said:

So is this the rumored standard AMD RDNA3 already supports before it was announced? 

Yeah, I remember some YouTube video talking about how AMD was going to have DP 2.1 and how that might be a typo because 2.0 was just announced. Looks like it's real after all. Not that surprising that a company implementing hardware might get earlier access to the specs.

[tikker]

Changing the DP8k one would've been good probably, it being so close to DP80. Seeing more USB4 integration is nice.

[Glenwing]

13 minutes ago, porina said:

Maybe I should add it somewhere as a feature request but I don't seem to be able to edit tables to add more cells in this forum and I don't have alternate software on my laptop to do it externally. I think it would be useful to see what resolutions are supported at what frequencies, with and without DSC.

 

DP and HDMI both have a similar problem. We know what port we have, but it isn't obvious what cable is needed for a given display. I suppose it is a safe choice to pick a cable that is at least as good as the display. In a similar way, when I was shopping for certified HDMI cables to go with my then new TV, I knew I needed a cable for HDMI 2.1. The correct name Ultra High Speed HDMI cable might as well not exist. A potential problem with using the standard number for the cable also is future standards may not advance requirements but version mismatch would be confusing. It seems like every naming decision in tech has problems one way or another!

I think if monitor manufacturers listed their ports as "1× DisplayPort input supporting HBR3 speed" or "2× HDMI input (40G FRL speed)" instead of "DP 1.4 port" and "HDMI 2.1 port", which theoretically is what they're supposed to do anyway, it would help a lot with this problem. Really it's just a matter of having matching information available between device and cable designations.

[Doobeedoo]

Neat extras rounding the spec though. Cable certification is a good thing. Longer cable support too. DSC support mandate is good, then again I'm much more interested to see new monitors with UHBR20 or 80Gbits without DSC use for like 4K 240Hz 10-bit for example. I'm excited to see such monitor with QD-OLED eventually.

The better integration of USB-C alt mode another useful thing really. Still hoping we see C starting to be everyone more.

 

But yeah nice, can't wait to see upcoming monitors with new DP though, huge bandwidth increase. Like I mentioned above, I want such monitor and we're close, we have 4K 240Hz LCD using DSC and we have QD-OLED monitor. Now to combine techs sans LCD and DSC and yey.