Display lag or something else?
-
ablemor
- Posts: 313
- Joined: 21 Oct 2022, 07:05
Re: Display lag or something else?
I have a 9800X3D and an X670E Hero, and I'm getting 80ns—nothing happens; RAM speed has nothing to do with this problem.
- kyube
- Posts: 991
- Joined: 29 Jan 2018, 12:03
Re: Display lag or something else?
List out your entire system & peripherals & what specific software settings you've changed.
This is insufficient information for any kind of troubleshooting.
0.) If you think "RAM Speed" has nothing to do with "this problem" (which is entirely based on HW choices users are doing and lack of knowledge in software tuning), then you're ignorant on how computer systems work.
1.) You're already severely lagging by using a ALC4080-based motherboard.
2.) "This problem" was never defined by anyone in these forums and it's all related to people being lazy & at this point, incapable to use basic tools to evaluate "the problem".
You have no idea how computer system performance works if you think chasing lower DRAM latency is irrelevant to decreasing total system latency or that it's somehow "impossible to tell".adrian_13371 wrote: ↑Today, 05:22You know that memory latency has nothing to do with peripheral input lag right? Even if it did you would not be able to tell because no human being is able to feel a 10 nanosecond difference
It's not +-10ns flat to the total system latency.
Every single cache miss (which is more prevalent on <32MiB CPUs & unoptimized games) requires your system to fetch data from DRAM, which is a expensive operation.
You can evaluate your L3 cache hit rate by using CapFrameX's new feature. You want to look at the complement value (miss rate = 1 - hit rate)
The miss rate is the amount of times in a specific time interval your CPU had to go through to DRAM or the worst case scenario, to your secondary/auxilliary memory (storage devices)
The deeper your CPU has to go, the higher the jitter in-game.
A 1ns difference is akin to 0,3m (30cm) in copper trace travel distance, as per Grace Hopper's analogy
1µs is akin to 300m in reducing copper trace travel distance.
It's a massive difference to your system.
Every single piece of software (OS, DRAM, games) relies on your DRAM R/W latency being as low as architecturally possible.
No, AIDA64 isn't a good way to evaluate DRAM R/W latency as it's a unloaded latency test.
Use Intel's Memory Latency Checker (IMLC, CLI version), microbenchmarksgui (clamchowder) or RamTestPro (RTP) Benchmark if you want a more valid test for both CPU platforms.
I still cannot believe that the moderation on these forums hasn't taken care of these type of FUD threads thus far.
evaluating xhci controller performance | audio latency discussion thread | "Why is LatencyMon not desirable to objectively measure DPC/ISR driver performance" | AM4 / AM5 system tuning considerations | latency-oriented HW considerations | “xhci hand-off” setting considerations | #1 tip for electricity-related topics | ESPORTS: Latency Perception, Temporal Ventriloquism & Horizon of Simultaneity | good lcd backlight strobing implementation list | display vs gpu scaling
