[Help] UHD, HEVC, and stuttering...

[deleted user]

Ok well, I think you definitely have some kind of cooling problem. Never worked with liquid cooling myself, so a few ideas:

1. Your mainboard (https://dlcdnets.asus.com/pub/ASUS/mb/LG...th_Z87.pdf) has a CPU_FAN and a CPU_OPT port, both of which seem to be available to power a CPU fan. So I would try to disconnect it from the one it's currently on (with PC turned off obviously) and put it into the other. For example, if it is in CPU_FAN right now, plug it into CPU_OPT (OPT stands for optional I believe) instead. See if anything changes (worth a try). Now, if your computer does not start after doing this, just reverse it. I read somewhere that some mainboards won't run the system if nothing is attached to CPU_FAN.

2. Check if all the water cooling tubes are arranged so that they aren't bent/prohibit water flow. Maybe generally check if there is still enough fluid in there and nothing is stuck/clogged and that the water circulation is running properly (not sure if that's something that can be a problem, as I never used water cooling as mentioned).

3. Otherwise I think it's safe to say either your cooling has a problem or your mainboard is malfunctioning and not regulating the cooling properly. If it's the first, maybe you still have warranty for the cooler? If it's the latter, not sure what can be done (maybe the guys on reddit can help).

4. If the cooler is the problem and you won't get a replacement with warranty, I'd just replace the cooler. You could also go get a cheap one just for trying if the problem is your mainboard, replacing the water cooling temporarily.

[spoRv]

The CPU_FAN is connected to the cooler fans; I tried to connect the cooler to CPU_OPT (and also other 3pin and 4pin connector) with no change.
Pipes were never touched, they are one cool and one warm as they should be.
Cooler is way long out of warranty... indeed, I also thought to buy a new cooler (or heatsink), but really, REALLy thinking to say WTF! (self censored) and buy a great replacement - and then sell mine for, more or less, what I paid for...

[deleted user]

Ah, bummer. Doesn't really tell us much unfortunately, as it's possible that CPU_FAN and CPU_OPT are operated by the same controller and if the controller is defective, they both would be affected.

Well, if it is defective, I don't think you'll be able to sell it too well.

At this point I recommend you do ask for help on reddit or similar, bc they may actually know the water cooling system you use etc. and give you real qualified answers instead of wild guesses (like myself).

If you prefer Discord, here's a technology Discord with a tech help channel: https://discord.gg/qRkPPQ

[spoRv]

I discorded! let's see what will happen!

I'll eventually reddit later...

[spoRv]

Today, PC decided to shut down during an (heavy) avisynth encoding... at the BIOS, changed fan speeds to "turbo", and system seemed more stable - indeed, previous encoding ended well.

So, I wanted to stress test a bit the system with the simple Cinegy Cinescore (that shut down yesterday topping temp more than 100°C )
It started well, with only sparsely temp peaks at 68°C, but the PC suddenly shut down nearly the end, with the temp around 62°C).
Now, adding this clue to the fact every connection I tried with the cooler to restore its LED light dimmed to full light (and consequently full cooling), I'm strongly supposing that the PSU is dying... anyone else agree?

[Colek]

It could be an issue, that PSU doesn't seem to be something very recommended brand. I'd try it with some Corsair/EVGA. You can use some of the wattage calculators online to see how much you need.

Also - you bought setup 2 years ago. Did you replace thermal grease under heatsink in that timespan?

[spoRv]

Not replaced thermal paste - according to some, it MUST be replaced every one or two years, while others say it should not be replaced, or depending by the quality/type of it.
But, as the PC shut down at 62°C, I'm mostly sure it's not an overheating problem, more something voltage related IMHO!

I also thought about the fact I'm using a powerstrip (economic one, I should admit!) since some time... may it be "guilty"?
The fact is, I have only one plug in my room, so I'll connect only the PC and see if the cooler LED is still dim - if not, hooray!

[deleted user]

Not replaced thermal paste - according to some, it MUST be replaced every one or two years, while others say it should not be replaced, or depending by the quality/type of it.
But, as the PC shut down at 62°C, I'm mostly sure it's not an overheating problem, more something voltage related IMHO!

I also thought about the fact I'm using a powerstrip (economic one, I should admit!) since some time... may it be "guilty"?
The fact is, I have only one plug in my room, so I'll connect only the PC and see if the cooler LED is still dim - if not, hooray!

Are you only watching the CPU temperature? Because there's also the chipset and other components on the mainboard that can overheat. GPU too by the way, but unlikely to happen just because you were encoding.

[spoRv]

All the other temps are at max 30°C - so, don't think they are overheating!

[deleted user]

Hmmm fair enough, I understand your suspicion then. Though, I would do some RAM test before going down more expensive routes. Spontaneous shutdowns could be RAM related I think. There are some RAM tests you can burn to a CD or put on an USB stick and boot from them and then let them run.