Latency in real world scenarios

alexkyriak · 14 Nov 2021 at 09:09

Morning all. Can anyone recommend a good article or video that explains the real world differences between a lower and higher latency, all other things considered? Ie, assuming MHz frequency is the same. There’s a bunch on the net but not sure I have found just a latency-focussed article/video. Am trying to learn more about this subject. Thanks

EsaT · 14 Nov 2021 at 09:59

Little short but, all those timings have effect this way:

In short while you want frequency to be high, you want timings to be low.

alexkyriak · 14 Nov 2021 at 11:48

Thank you mate, that’s useful. But what does a small improvement in the nanoseconds actually translate into though? How would I see the improvement in gaming? I assume FPS improvements? I don’t have a feel for the benefit from say 0.2ns difference, or 1ns or 10ns, would be in a tangible sense. Not sure if I’m asking the question correctly!

MagicBoy · 14 Nov 2021 at 11:55

It depends.

There's no generalised answer. Some games won't show any difference, some games show a detectable increase.

HaloUK · 20 Nov 2021 at 09:15

Thanks for posting that @EsaT , very helpful. Nice one!

Guest2 · 25 Nov 2021 at 15:40

alexkyriak said:
Thank you mate, that’s useful. But what does a small improvement in the nanoseconds actually translate into though? How would I see the improvement in gaming? I assume FPS improvements? I don’t have a feel for the benefit from say 0.2ns difference, or 1ns or 10ns, would be in a tangible sense. Not sure if I’m asking the question correctly!

I saw quite big FPS gains (when a game had high fps) going from XMP 3600Mhz CAS16 to 3800Mhz CAS14.
The main improvement for me was the desktop performance seems to shoot up - clicking around in start menu, opening files and folders, context menus etc. It went from a very small delay on XMP to almost feeling instant (much smaller delay) on lowered timings. This was going from around 60ns memory latency to 53.3ns