Can DisplayPort 1.2 support 10-bit (8+2FRC) color at 1440p and 144hz?

Question

Generally speaking, I'm trying to determine if a given DisplayPort version will support difference display modes (resolution, refresh, and color depth). For this specific instance, I'm looking at the MSI Optix MAG274QRF-QD 27" monitor. I found a useful chart here that shows the bandwidth available to given DisplayPort versions - and according to it DisplayPort 1.2 supports 17.28gbps. Using some napkin math, a 10-bit, 1440p@144hz signal would consume:

10 bits * 3 colors * 2560 * 1440 * 144 = 15.925248 gigabits per second

By this measure this scenario should be fine - and for what it's worth I think 8+2 bit color actually only consumes 8 bits per color channel of bandwidth, not ten. However, in a review by rtings.com they mention specifically: "The max refresh rate you can achieve with a 1440p resolution and a 10-bit signal over DisplayPort is 120Hz..."

Not only does this disagree with the math above, it also doesn't make sense from a marketing perspective why this monitor would be advertised with these features but unable to use them. I checked the manual for the monitor itself and it lists 1440p@165hz as a viable resolution - but it does not mention color-depth.

Does DisplayPort 1.2 support 1440p@144hz with 10-bit color depth?

Answer 1

Of course it makes sense to advertise all the features, even if they are not supported simultaneously - they sell more displays this way. For example TVs are advertised do 120 Hz and 8K, but they can't do them simultaneously, only 4k@120 or 8k@60.

The manual is very vague, so is the datasheet. There is no clear answer from the manufacturer.

And your calculations assume there is only raw pixel data sent, which is not true.

According to Wikipedia DisplayPort article here, 1440p@144 Hz with 10-bit color depth needs 17.60 Gbit/s, which the DisplayPort 1.2a can't support at standard CVT-R2 signal timing, but if both the display and video card supports it, tweaking the video signal timing parameters could theoretically work.

For comparison, 1440p@120 Hz only needs 14.49 Gbit/s which should work just fine.

Answer 2

8+2FRC is an LCD panel trait, it doesn't imply that the other parts of the display chain are required to use 10 bit depth. If any it means that such a signal would not go inherently to waste.

But for most intents and purposes deep colour would only really be notable with HDR (which isn't exactly going to be great on a low contrast IPS screen). And if your gpu can do 8-bit with dithering, that advantage is almost academical.

That's probably the reason why MSI didn't feel like complicating the product page on their website with too many ifs and buts (if they even realized that could be a potential use case to begin with), or even the monitor's EDID with any step in-between 165 and 120 hertz (which they did instead for HDMI, given 144Hz is the highest common frequency version 2.0 could support without chroma subsampling).

With this said, Displayport 1.2 (HBR2) should be more than enough for 2560x1440@144Hz at 10 uncompressed bpc (in fact your very monitor was reported to sustain even 150Hz in such conditions). It's just that you need slightly tighter timings than the CVT-RBv2 standard would otherwise instruct (my LG 27GN850 and others do exactly this out of the box for instance).

And especially this aspect is a sore point of back-of-the-envelope calculations all around the net. As can be seen in the tool linked in the comments or here, "active pixels data" doesn't just happen to travel alone in a vacuum. Although the terminology may sound incredibly geriatric, even the most modern circuitry still requires some sort of additional "delimiting" information between frames.

How it is calculated is a tedious story for another day, but TL;DR you have to do your math against the whole transmitted signal inclusive of the blanking interval, not just the "visible" resolution.