It's as if the cooler does nothing at all?

[Solrosen]

CPU: i7 7700k Kaby Lake (OC'd to 4.7ghz)

Cooler: EVGA CLC 280

OLD fans: Corsair Air Series AF140 Quiet edition

NEW fans: Arctic P14 (PWM PST)

New thermal paste: Thermal Grizzly Kryonaut

Motherboard: STRIX Z270E Gaming

 

Been noticing high temps in my CPU for some time (41 at idle, around 80-90 during high loads) decided to upgrade to high pressure fans with better thermal paste.

 

Here are some interesting observations I've done after changing them.

1. Idle temps are a lot better, average at 33 degrees when it's stable (could be the paste maybe?).

2. Temps fluctuate a whole lot when doing everyday desktop or idle work, jumps randomly and rapidly between 33-75 degrees, maybe this is normal but keep reading.

3. I tried benching it with cinebench and noticed barely no decrease in temps, maxes out at about 85-90 degrees.

4. HW-monitor says the CPU-fans are constantly spinning at ~1500rpm, however when I open the EVGA software "flow control" it says something else. I might have connected to the wrong header?

5. With flow control I can set the pump and fans however I want, and if I MAX them out, do the bench, there is no difference in temps, 85-90. If I lower the pump and fans to it's lowest setting then run the test it's the same story, no change in temps, 85-90.

6. If i sit in a practice map on CSGO, with several bots slashing their knifes at me, with fans set to manual on the lowest rpm (~400) the CPU sits at about 70 degrees. If I then max the fans (~1700rpm) nothing happens to the temps, stays at 70.

 

I made this in a rush so it might be poorly formulated.

NOTE, I know this is a CPU meant for overclocking, so getting close to 100 degrees celsius isn't necessarily scary.

 

So what's going on? Is the cooler not functioning properly? Was it me that installed the paste and cooler itself poorly? Am I getting inaccurate readings?

I'm going to disassemble it tomorrow and look at the spread of the paste to see if something has gone awry.

 

Included image shows what the temp fluctuations looks like whilst only typing this post.

 

Please, if you have any ideas feel free to share, even if it's just "might be weird, but you're within acceptable temps so 'fuggedaboud it kid' ". It is still noteworthy.

[Guest]

How much did you increase the voltage while overclocking?

[Somerandomtechyboi]

14 minutes ago, Solrosen said:

I know this is a CPU meant for overclocking, so getting close to 100 degrees celsius isn't necessarily scary.

At a low volt like 1.3v or under its not scary but above 1.4 then it will become quite a problem, above 1.4 its pref to be kept under 90-85c just to make oc abit more stable and less degradation

 

Though for 4.7 you are prob only running a volt like 1.3 so 100c is not an issue just abit uncomfortable

 

16 minutes ago, Solrosen said:

With flow control I can set the pump and fans however I want, and if I MAX them out, do the bench, there is no difference in temps, 85-90. If I lower the pump and fans to it's lowest setting then run the test it's the same story, no change in temps, 85-90.

At this point im suspecting some of the water has permeated and/or the pump is dying so no sht why temps are rather eh, consider replacing the aio

[Somerandomtechyboi]

17 minutes ago, Caroline said:

I believe you both forgot to add an /s after those statements.

Wdym by /s?

[Solrosen]

1 hour ago, WickedThunder86 said:

How much did you increase the voltage while overclocking?

Honestly haven't touched my CPU settings in a loooong while so I checked, according to my BIOS the core clock is at 1.168v at the moment

Scratch that, according to HWM it maxes out at about 1.360V when under load like a cinebench test.

 

Please excuse my bare knowledge of these sorts of things, I'm not exactly well read on PC hardware.

[Solrosen]

52 minutes ago, Caroline said:

I believe you both forgot to add an /s after those statements.

 

Ah... anyway OP.

You do know those CPUs have mayo for TIM inside right? and that they're prone to overheating, temp spikes and throttling because of that? Running the pump and fans faster won't do anything because the problem lies between the die and IHS. You should start considering delid if you're overclocking that high. Temps will probably drop by 10C overall with paste (NOT liquid metal) if vcore stays below 1.3, SA and IO under 1.15 if BCLK and/or memory are overclocked as well

 

Imma be honest and just say that I'm unfamiliar with most of the terms you just used. Speak "amateur language" for me if you could

[WoodenMarker]

6 hours ago, Solrosen said:

according to HWM it maxes out at about 1.360V when under load like a cinebench test. 

That's high voltage and may explain your temps. 

If you're going to oc, set your voltage manually. 4.7ghz should be easily done under 1.3v and its best to use as low voltage as possible while maintaining stability. 

[thrasher_565]

what is the case? a pic of the setup would be ideal instead of asking a milling questions.

[Guest]

11 hours ago, Somerandomtechyboi said:

Wdym by /s?

/s means everything you said was sarcasm.

[Somerandomtechyboi]

20 minutes ago, WickedThunder86 said:

/s means everything you said was sarcasm.

 

12 hours ago, Somerandomtechyboi said:

At a low volt like 1.3v or under its not scary but above 1.4 then it will become quite a problem, above 1.4 its pref to be kept under 90-85c just to make oc abit more stable and less degradation

 

Though for 4.7 you are prob only running a volt like 1.3 so 100c is not an issue just abit uncomfortable

Not saying this sarcastically

 

I abuse my hardware quite alot and most of em havent degraded yet, even to the point of running a sacrificial pentium to 100c at i think ~1.75v, still alive and hasnt seemd to degrade yet, def doing something like this will degrade the crap out of it overtime due to high volt and temps but running like 1.3v at 100c i dont really see an issue with that, its just a matter of personal preference

 

I dont see an issue with running over 100c when you are using low volts like 1.3 nor do i see an issue with running cpus near absolute max "safe" volts (0.3v under deathzone), if you actually take care of your cpus then you wouldnt abuse em like me

 

but theres a fine line between abusing the sht out of your hardware and having it degrade and having the hardware actually die or babying your hardware too much and ruining your performance

 

I abuse my hardware if it isnt evident already (running 2v through a 45nm cpu quite a few times, running cpus over 100c with high voltages, etc.), so take my advice with a grain of salt xD