ffs_ wrote: ↑01 Nov 2020, 02:31
Apparently 50 us was about latency while idle and 500 us was about latency while playing the game. But idle latency is rather a spherical chicken in vacuum. If you have bad latency during idle, of course you need to lower it. But you could have good latency during idle and huge spikes (1000+ us) during playing the game. So check idle latency AND latency while playing. 5-10 or 50 us during idle doesn't mean much if you get 500+ us or even 1000+ us during playing the game. And if you get 200-300 us while playing the game anyway, it doesn't really matter if using HPET raises idle latency, and it doesn't mean that your
input-lag was raised as well. That was the point.
And you're still missing the point that the idle latency is generally the quickest the system can respond (assuming outside of a power saving mode). A system under load should have even higher input lag assuming any of the CPUs or resources handling the mouse/kb/GPU interrupts are affected (again there are other factors to consider, e.g. manually assigned interrupt affinity to combat this, how much load is applied).
HPET overhead applies to both idle and load, and hence can increase latency in both cases as well, thus affecting input lag. Whether or not you notice it and/or prefer it on/off in your games or programs is a different thing.
If you are so adamant that HPET:On + useplatformclock:true does not increase input lag at all, simply put it to the test for your case.
Either use an LDAT if you have access to one, reflex analyzer, or very high FPS capture with all the correct camera settings like how Chief does it.
ffs_ wrote: ↑01 Nov 2020, 02:31
howiec wrote: ↑31 Oct 2020, 17:43
Duh, I mentioned Intel in response to your post to be honest that I haven't tried Ryzen. Then you say that's why you mentioned Ryzen. Then I mentioned it should be platform independent (as far as increased overhead is concerned). Then again... circles. Let's not debate the irrelevant.
Nope, I'm talking about your
original post where you gave a
general advice:
howiec wrote: ↑27 Oct 2020, 14:08
However, as he mentions, do
not enable HPET if you want low latency. I highly recommend to ensure it is disabled for most gamers.
after which I made a gentle hint that results might be different for different platforms. if you mentioned from the beginning that your advice is applicable to Intel platform mostly and you're not sure about AMD, I wouldn't say a word.
Why are we still talking about this?
I gave general advice because HPET has increased overhead by design.
Then you mentioned your Ryzen system.
Then I mentioned Intel to be clear that I haven't tested Ryzen firsthand.
Then you said that's why you mentioned Ryzen.
We're saying the same thing: We're on different platforms...
Yes, I can say a word because simply looking at the actual HPET architecture vs TSC, it's obvious that it's slower/more costly to access.
The performance loss may not be as bad on a newer Ryzen vs current Intel systems but it still cannot be as "fast" as TSC.
There are various test and supporting data out there. Heck, even the Linux kernel code bashes HPET to some degree for various related and unrelated reasons.
https://github.com/torvalds/linux/blob/ ... nel/hpet.c