how i fixed the problem.

[dervu]

Complete offtopic (or maybe not) - IIRC you had 5600x and some b550, asus or msi tomahawk one? You swapped these for 3d & crosshair viii, respectively? Got any improvement? Or how does it feel at least (5600x vs 3d cpu)? My gear is almost the same as yours, that's why I'm asking.

I did not replace only mobo and CPU. I replaced everything:
- PSU
- Mobo
- CPU
- GPU
- RAM
- Drives
- Keyboard
- Mouse
- Headphones
- Monitor
- Routers
Even replaced mousepad to skypad, to rule out muddy mousepad factor to be sure in oncoming months.

Only ethernet cable and displayport cable are old.

I turned out everything was broken, got most difference from switching mobo, but every other part also had its share.
If I had to evaluate which part was most affecting that issue would be:
Motherboard -> M2 SSD -> Mouse -> Headphones -> Keyboard -> Routers.

Also on B550 Gaming Plus from MSI I had desync in timers in CPU-Z (version after MSI bug fix) that I could see even after replacing it with same new model (it was only good at first start), but it was before I replaced M2 SSD. With new motherboard from Asus (X570) and new M2 SSD it did not happen.
That is why I switched to X570, because it was weird for me that it would still happen on B550, even though someone from MSI forums with MSI board checked it for me and it did not happen for him at all, so something was wrong. Did not want to risk going with same mobo again. Maybe that M2 drive was somehow affecting something, idk.

Still, I would not say that is fixed, even if it is for now, because to say that I would have to play at least half a year to be sure it doesn't come back and I had too much false positives in recent years with trying different "fixes".
Toroids on power cables, running on online UPS battery, isolation transformer, using capacitors used in capacitor banks, dc blockers.

What is interesting is that running on UPS battery and not having improvement could paradoxically mean two things:
- Issue is in PC - like real hardware issue not originating from electric grid, because you rule out electric grid - no real improvement (assuming that nothing with battery is wrong at that moment)
- Issue is in electric grid - PC is broken - that is why you don't get real improvement

At least it works like that in my case.

It is safest to replace everything if you have at least some guess where source of the problem is (when difference is huge after disconnecting faulty device or whatever else was done). Remove that source and replace whole setup.
If you should replace your setup also depends if your PC works flawlessly in another location. To be completely sure, you would have to use computer at that another location that never ever had this issue and compare your PC to this other PC at that other location.
All comes down to experience and knowing what real gameplay should look like.

Still even after fighting with this crap for years I can't get real proof like everyone else can't. Maybe someday I will have to do another 1000fps cam recording and try to compare that to older ones with 1:1 setup, but so far my experience was it is really hard to find anything. While you can tell immediately something is wrong when using PC affected by this issue.

Even if I wanted to recreate this issue with device and its faulty connections that was affecting it all, it is too risky for me financially (risk of new PC getting damaged too) and would require tons of money and free time.

[triplese]

What is interesting is that running on UPS battery and not having improvement could paradoxically mean two things:
- Issue is in PC - like real hardware issue not originating from electric grid, because you rule out electric grid - no real improvement (assuming that nothing with battery is wrong at that moment)
- Issue is in electric grid - PC is broken - that is why you don't get real improvement

If electricity was not turned off completely with breaker, you still getting EMI source in your wall - wires can be "antennas". And even if you turned it off - you have EMI source in wall from your neighbors, sharing wall.
To make 100% sure its electricity (or not) related, someone need just to take oscilloscope and spectrum analyzer and take measurements in "good" and "bad" places and then compare it. I think in Europe its can be easily done by renting equip or calling qualified electrician laboratory if you dont have degree in metrology and standardization, just google it. Private electrician lab giving all measurements and protocol with conclusions, so it will be big and definite answer to this question. First who will do this can start youtube channel and film it and have big profit, because such investigation will be viewed better then Netflix movies And its not costs too much, its definitely smaller price than buying new PC.
So if graphs and spectrum correlation is low - it can be EMI/electricity. If correlation is high (or graphs almost equal) - its not electricity related.

[dervu]

What is interesting is that running on UPS battery and not having improvement could paradoxically mean two things:
- Issue is in PC - like real hardware issue not originating from electric grid, because you rule out electric grid - no real improvement (assuming that nothing with battery is wrong at that moment)
- Issue is in electric grid - PC is broken - that is why you don't get real improvement

If electricity was not turned off completely with breaker, you still getting EMI source in your wall - wires can be "antennas". And even if you turned it off - you have EMI source in wall from your neighbors, sharing wall.
To make 100% sure its electricity (or not) related, someone need just to take oscilloscope and spectrum analyzer and take measurements in "good" and "bad" places and then compare it. I think in Europe its can be easily done by renting equip or calling qualified electrician laboratory if you dont have degree in metrology and standardization, just google it. Private electrician lab giving all measurements and protocol with conclusions, so it will be big and definite answer to this question. First who will do this can start youtube channel and film it and have big profit, because such investigation will be viewed better then Netflix movies And its not costs too much, its definitely smaller price than buying new PC.
So if graphs and spectrum correlation is low - it can be EMI/electricity. If correlation is high (or graphs almost equal) - its not electricity related.

I also did that test months ago, there was no difference.
Also specialists from regulatory institution did not see anything that according to them could affect PCs in non-conducted interferences using professional spectrum analyzer. They told me if anything like that was affecting PC, it would have to be really really close and really strong. There is nothing like that here. Also tried researching using my own RFI spectrum analyzer, none of signals are extraordinary strong.

I am not sure if those EMC labs will do supraharmonics measurements, but there are some devices that can do that, so who knows.
Even if you find source and it is outside of your apartment it becomes really hard to turn source of interference off, because that part of spectrum might not be regulated at all, so noone will care.
You can wait for it to disappear, but guess what, you will have to pay EMC labs shitton of money to wait for it too.
The problem is that, you have to correlate those things and have something more than feelings about how PC works and why, so you need at least:
- Repeatable mouse moving device powered from separate power source
- 1000fps camera
Maybe that would give some views.

Maybe someday if I will be sure for months that it doesn't come back, I could replicate that, but still that is big $ risk, but also big $ reward.

[Cyanide_]

Complete offtopic (or maybe not) - IIRC you had 5600x and some b550, asus or msi tomahawk one? You swapped these for 3d & crosshair viii, respectively? Got any improvement? Or how does it feel at least (5600x vs 3d cpu)? My gear is almost the same as yours, that's why I'm asking.

I did not replace only mobo and CPU. I replaced everything:
- PSU
- Mobo
- CPU
- GPU
- RAM
- Drives
- Keyboard
- Mouse
- Headphones
- Monitor
- Routers
Even replaced mousepad to skypad, to rule out muddy mousepad factor to be sure in oncoming months.

Only ethernet cable and displayport cable are old.

I turned out everything was broken, got most difference from switching mobo, but every other part also had its share.
If I had to evaluate which part was most affecting that issue would be:
Motherboard -> M2 SSD -> Mouse -> Headphones -> Keyboard -> Routers.

Also on B550 Gaming Plus from MSI I had desync in timers in CPU-Z (version after MSI bug fix) that I could see even after replacing it with same new model (it was only good at first start), but it was before I replaced M2 SSD. With new motherboard from Asus (X570) and new M2 SSD it did not happen.
That is why I switched to X570, because it was weird for me that it would still happen on B550, even though someone from MSI forums with MSI board checked it for me and it did not happen for him at all, so something was wrong. Did not want to risk going with same mobo again. Maybe that M2 drive was somehow affecting something, idk.

Still, I would not say that is fixed, even if it is for now, because to say that I would have to play at least half a year to be sure it doesn't come back and I had too much false positives in recent years with trying different "fixes".
Toroids on power cables, running on online UPS battery, isolation transformer, using capacitors used in capacitor banks, dc blockers.

What is interesting is that running on UPS battery and not having improvement could paradoxically mean two things:
- Issue is in PC - like real hardware issue not originating from electric grid, because you rule out electric grid - no real improvement (assuming that nothing with battery is wrong at that moment)
- Issue is in electric grid - PC is broken - that is why you don't get real improvement

At least it works like that in my case.

It is safest to replace everything if you have at least some guess where source of the problem is (when difference is huge after disconnecting faulty device or whatever else was done). Remove that source and replace whole setup.
If you should replace your setup also depends if your PC works flawlessly in another location. To be completely sure, you would have to use computer at that another location that never ever had this issue and compare your PC to this other PC at that other location.
All comes down to experience and knowing what real gameplay should look like.

Still even after fighting with this crap for years I can't get real proof like everyone else can't. Maybe someday I will have to do another 1000fps cam recording and try to compare that to older ones with 1:1 setup, but so far my experience was it is really hard to find anything. While you can tell immediately something is wrong when using PC affected by this issue.

Even if I wanted to recreate this issue with device and its faulty connections that was affecting it all, it is too risky for me financially (risk of new PC getting damaged too) and would require tons of money and free time.

Thank you for your comprehensive response! Yeah I've read about importance of swapping MoBo, I've also read about PSU, RAM swapping and guys reported that it worked. I did not believe in that "infected PC" theory but after moving houses and getting literally the same lag after time, I'm thinking of it so maybe PSU/MoBo are somewhat damaged permanently and should be swapped.

[triplese]

I am not sure if those EMC labs will do supraharmonics measurements, but there are some devices that can do that, so who knows.
Even if you find source and it is outside of your apartment it becomes really hard to turn source of interference off, because that part of spectrum might not be regulated at all, so noone will care.
You can wait for it to disappear, but guess what, you will have to pay EMC labs shitton of money to wait for it too.
The problem is that, you have to correlate those things and have something more than feelings about how PC works and why, so you need at least:
- Repeatable mouse moving device powered from separate power source
- 1000fps camera
Maybe that would give some views.

Maybe someday if I will be sure for months that it doesn't come back, I could replicate that, but still that is big $ risk, but also big $ reward.

You can try to use something like EEM-MA770, its measuring harmonics. Its price is about 500 euros, and you can return it if do measurements relatively fast.
If you find RFI outside your apartments, you can claim it to your regulatory, if you have protocol. There are some successful precedents with powerline ethernet adaptors, energy providers force users to turn it off. So community needs just to find root cause and then it will be fixed really fast - journalists, esports players will make so many whine about it...
Mouse moving device can be easily done with arduino, solenoid and few bricks of LEGO and lifepo4 battery. I saw this devices on youtube testing input lag of mice, but cant find it right now.