Question Calculating gpu to monitor capabilities of rtx 2080

[Zimrathon]

Hi, new to the finer details of gfx card tech specs so sorry if this turns out to be a stupid question, but how do you work out the max capability of a gpu applied to a monitor/monitors set up.

For example, the way I see it, from my layman’s view, if I have 3 monitors at 1920 x 1080 resolution and a colour depth of 24bits, the gpu needs to shift 3 x 1920 x 1080 x 4 (colour depth) bytes per single ‘frame’ across all 3 monitors = 24,883,200 bytes. If, as per nvidia site, the 2080 has 448GB/s bandwidth, it means that it could output 448GB / 25MB ‘frames’ per second, which is around 18,000. Clearly I’m doing something wrong, can any expert point out my error please.

Thanks

 

[kanewolf]

Memory bandwidth isn't directly related to output. The memory has to do other things at the same time.
And each interface port has specific bandwidth limitations. The display port 1.4 and HDMI 2.0b each have bandwidth limitations. There are lots of limiting factors in display.

 

[Zimrathon]

Memory bandwidth isn't directly related to output. The memory has to do other things at the same time.
And each interface port has specific bandwidth limitations. The display port 1.4 and HDMI 2.0b each have bandwidth limitations. There are lots of limiting factors in display.

Thanks for your answer. How does one then determine what spec of monitors will make best use without overspeccing. For example, on my rtx2080 example, how can I know whether it will support 3 1440 monitors @ 144hz?

Thanks

 

[kanewolf]

Thanks for your answer. How does one then determine what spec of monitors will make best use without overspeccing. For example, on my rtx2080 example, how can I know whether it will support 3 1440 monitors @ 144hz?

Thanks

You would have to look at the specs for the card. Each interface will have a version (HDMI 2.0b for example). You can then look up the definition of HDMI 2.0b and see what display resolution and refresh that interface can support. You could do that for each interface. Then the card will have a maximum resolution the card can support documented in the card specifications. BUT being able to support X pixels by Y pixels at 60hz doesn't tell you anything about whether you could game at that resolution. So if you were trying to do spreadsheet work you could look at specs. For realtime displays like games you have to look at benchmarks and other user's experiences with the specific software you are using.

 

[Zimrathon]

You would have to look at the specs for the card. Each interface will have a version (HDMI 2.0b for example). You can then look up the definition of HDMI 2.0b and see what display resolution and refresh that interface can support. You could do that for each interface. Then the card will have a maximum resolution the card can support documented in the card specifications. BUT being able to support X pixels by Y pixels at 60hz doesn't tell you anything about whether you could game at that resolution. So if you were trying to do spreadsheet work you could look at specs. For realtime displays like games you have to look at benchmarks and other user's experiences with the specific software you are using.

Sorry, I’m familiar with the above advice but thanks for taking the time to share. I was more meaning the combination of the three monitors being able to be kept pace with.

I understand that for each output port I could check the spec of the tech to know whether it can output, for example, 144hz @ 2k, but, if in doing so I find it can theoretically do so on three separate ports, how do I check that the gpu is capable of pushing out to all three in parallel, satisfying three monitors, if there are the said overheads to consider, as in the first reply, and I can’t simply use nvidias stated overall throughput of 448GB.

Thanks

 

[kanewolf]

Sorry, I’m familiar with the above advice but thanks for taking the time to share. I was more meaning the combination of the three monitors being able to be kept pace with.

I understand that for each output port I could check the spec of the tech to know whether it can output, for example, 144hz @ 2k, but, if in doing so I find it can theoretically do so on three separate ports, how do I check that the gpu is capable of pushing out to all three in parallel, satisfying three monitors, if there are the said overheads to consider, as in the first reply, and I can’t simply use nvidias stated overall throughput of 448GB.

Thanks

I believe the only answer is benchmarks done in reviews or other user experiences with the exact software you are interested in using. There is no "paper" method because the efficiency of the software and the capabilities of the rest of the system has to be factored in.