[SOLVED] I9-9900K Idle temps

bchen_16

Prominent
Jul 7, 2019
14
0
510
I'm getting around 40c idle temps on my 9900k with a h150i pro cooler, is this normal? I haven't done any overclocking and everything in the bios is default except xmp. I'm using the stock paste that came with the AIO.
CPU: i9 9900k
Cooler: Corsair H150i Pro
Motherboard: Asus Maximus Xi Code
Ram: 16GB Vengeance RGB Pro
SSD/HDD: WD SN750 500GB/Seagate Barracuda
GPU: ROG Strix 2080 OC
PSU: EVGA SuperNOVA G3 80+ Gold
Chassis: Lian-li PC-O11
Fan Configs: 3 ML120 Intake/6 LL120 Exhuast/ Top mounted rad with LL120
 
Last edited:
Solution
Depending on your ambient temperature, yeah, it might be. It's at the high range of where we like to see idle temps, which don't matter much EXCEPT when they are too high because it generally indicates a problem with the cooler mounting or excessive voltage, but for the 9900k it's not too unusual. It is probably SOMEWHAT unusual given the cooler you have though.

You might want to increase the baseline speed of your radiator and case fans slightly, maybe about an additional 5% PWM signal at 40°C or lower on your fan curve to see if that drops your idle temp into the 35° range or not.

Also, exhaust configuration isn't usually the best for cooling performance. My recommendation would be to move the radiator to the front as an intake...
Depending on your ambient temperature, yeah, it might be. It's at the high range of where we like to see idle temps, which don't matter much EXCEPT when they are too high because it generally indicates a problem with the cooler mounting or excessive voltage, but for the 9900k it's not too unusual. It is probably SOMEWHAT unusual given the cooler you have though.

You might want to increase the baseline speed of your radiator and case fans slightly, maybe about an additional 5% PWM signal at 40°C or lower on your fan curve to see if that drops your idle temp into the 35° range or not.

Also, exhaust configuration isn't usually the best for cooling performance. My recommendation would be to move the radiator to the front as an intake configuration in order to take advantage of the much cooler ambient air, rather than trying to cool it with hot air coming directly off your motherboard, CPU and graphics card as well as whatever residual heat get added by the drives and PSU via radiant contributions.
 
Solution
Thanks for the response.

My ambient is 24-25c and it seems to idle at around 40-41 and rarely dropping below that. I do have the fans on the rad and the intake fixed at 1000RPM, exhuast at 800RPM, and the pump at around 2000RPM and the liquid temp would be around 32C.

Even if a crank every fan in the case to maximum, the temp just wouldn't stay in the high 30s, it would dip to 39 and immediately come back up to the 40s.

Also I do realize I have a negative air pressure situation, but for the sake of aesthetics I kept it like this.

IMO I feel like it has something to do with the thermal paste, because when I installed the pump I think i lifted it up after making contact with the IHS because the screws weren't aligned. Should I use something like TG Kryonaut? Here's my fan config
Bottom: 3X ML120
Side: 3X LL120
Top: 3X LL120 with the rad on top
 
Negative air pressure IMPROVES cooling peformance, so I have no issues with that. More dust, but better cooling. No concerns with that part.

My only thought is that I know for a fact that I've worked on many threads (And IIRC at least one moderator has specifically tested the theory on their own system with positive results) where somebody had a similar configuration as yours, with similar cooling issues, and moved their radiators to the front and resolved the entire problem due to the source air being ambient rather than already warmed inside the case.

Honestly though, idle temperatures really don't matter much so long as your maximum temperatures under a full TDP load are within tolerance and specifications, which for your system would technically be 80°C. A little less, better, but anything at or below 80 is acceptable.

Kryonaut is very good. It's what I use almost exclusively these days. Kryonaut, Noctua NT-H1/H2, Cryopaste, for high end machines and Arctic silver 5 or MX4 for mainstream machines.

It couldn't hurt to repaste if you think you botched the paste job. Be sure to clean everything well with isopropyl alcohol.

I'd worry more about maximum temps though. Run Prime95, choose the small FFT option and DISABLE all of the options for AVX at the bottom of the test selection screen. AVX2 and AVX should be disabled when doing thermal testing. Run for 15 minutes. If you exceed 80°C there is an issue, although these do run very hot so it won't be that surprising, but you also have a pretty capable cooler. You might consider better fans on your radiator as well. Those are ok, but there are much better options out there.
 
  • Like
Reactions: bchen_16
I don't have the computer with me right now, when I get home I will run Prime 95 and post the results after. When gaming for couple of hours, it caps out around 73c but I'll post the results later.
 
So I ran the tests exactly like you suggested. The package capped out at 78c with the fans on 1000RPM and pump around 2000. The package took 140 watts at max. And around 1.27-1.28 volts on all cores.
 
Are you putting a static speed on the fans? I'd let them run full range from around 800rpm at temps below 40°C to around 1000rpm at about 55°C and then full speed at about 75°C.

Regardless though, that's clearly within spec. You could get that temp a bit lower if necessary, but that is acceptable the way it is. Definitely within spec.
 
Ok great, I keep the fans at low RPM because of noise mainly. I changed the fan configuration to push pull on the radiator and the temps dropped around 2-3c on idle. Regardless I'm going to get some TG Kryonaut on Prime day just to get it all done at once. Thanks for the help!
 
After a couple of days of doing some digging, it seems like more of a software problem vs hardware.

I disable the iasync.exe (something like that) process and the temps dropped dramatically. The software has something to do with Intel Rapid Storage Technology.

I realized that it was IRST that was putting load on the CPU so I tried disabling it in the BIOS, however it just led me to a blue screen on boot. So I terminated the process in Task Manager forever and everything is running fine now.

How I found this was I saw core 1 was overloaded at 90% on idle and since it's a brand new computer there probably isn't a virus. So I went into the task manager and the rest is history.