I have noticed, after disabling all those my latencies have improved a lot more. Ill post a screenshot.
Wdf was around 300-500 before, ntos would jump to 1000's. Now it seems much better for me. And the current measured interrupt to process latency is around 10-20 now, with occasional spikes to 100-200.
I always had latency issues since i built my amd system, so this has been an obsession of mine. I also see people on amd forums, having same issues with audio pops and crackling on official amd forums with high end 7000 series cpus.
Yeah.........no. Latencymon is not a benchmark and is unpredictable as hell. Your 3 minutes and 50 seconds of recording means next to nothing. 200-300us is well within margin of error at idle.
If I leave it on while just doing my normal stuff on desktop, it'll regularly catch 1000us+ spikes that are meaningless and cause no adverse effects whatsoever. Conversely, trying to use Latencymon to track down visible and reproducible in-game stutters often doesn't turn up anything, only in the absolute worst cases of regular 10000us+ latency where something is seriously wrong. I have had the Zen 3 audio issues before on past CPUs/boards, and I have had specific in-game stutters. LatencyMon couldn't find anything.
Maybe pick a thread and stick to it (preferably your original, own thread on latencymon stutters), instead of asking the same thing in 3 different places.
LatencyMon is not a performance monitor. It's a self-described "audio suitability" checker, and its usefulness even in that role is pretty suspect. It's worth a quick look if you're having stutters, but it's not an incredibly useful tool.
Another big factor was the windows 11 security settings, they keep disturbing all the windows drivers, including, gpu, audio and network drivers. Denuvo and easy anticheat on some games.
I would like to discuss more, please give what i did a try and report your results.
I don't use VBS in Win11 because I run TPM disabled. That's really the only thing I can think of that makes a consistently appreciable and repeatable difference (penalty) to performance.
Denuvo and Easy Anticheat are not something the hardware can control.
Like I said, I run CPPC On and Preferred Cores Off because that's the 5800X3D recommendation. You can experiment with Preferred Cores to see if it makes a difference. There's no reason to disable CPPC.
Again, Latencymon is not a benchmark. Until you can show a clear difference in benchmarked game performance, sitting there staring at Latencymon is a huge waste of time.
