Jump to content

I9 9900K RUNNING AT 100°C under load (stock)

st4rway
 Share
Posted

hi!

 

Been using this PC mostly for work and, after running into some PC freezes while playing HL Alyx i decided to find the root of the issue and did an prime95 cpu stress test.

while having the cpu at stock, with no OC whatsoever, i've reached package temps of 100°C wich are way too high for stock frequency.

Other than a possible not so good airflow issue i have absolute no idea whats going on and what to do to solve it.

 

here's my configuration:

Mobo: Gigabyte Z390 Master

CPU: i9 9900K

RAM: 64GB - 4 x 16gb DDR4 Trident z 3000

GPU: MSI gaming x trio RTX 2080ti

PSU: Corsair RMx 1000W

M2 1: Samsung Evo 960 250gb

M2 2: Crucial P1 1 tb

M2 3: 660p 2TB

AIO cooler: Thermaltake Ring 360 (with 3 120 fans)

other fans: 3 x 120 thermaltake ring + 3 x low profile fans 

Case: Lian Li O11Dynamic

 

(Self assembled, Pea method on thermalpaste)

 

Idle temps: 36°C to 40°C

Room temp: 20°C

 

Airflow diagram and fan curve profiles in the attachment

 

Please help :)

 

 

 

airflow.jpg

fancurve.png

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Did you apply enough thermal paste or added enough pressure to the mounting for the cpu block? Also, did you remove the sticker attached to the mounting block? Ive seen a few nincompoops forgetting to remove that lol

Im with the mentaility of "IF IM NOT SURE IF ITS ENOUGH COOLING, GO OVERKILL"

 

CURRENT PC SPECS    

CPU             : Ryzen 5 3600 (Formerly Ryzen 3 1200)

GPU             : ASUS RX 580 Dual OC (Formerly ASUS GTX 1060 but it got corroded for some odd reasons)

GPU COOOER      : ID Cooling Frostflow 120 VGA (Stock cooler overheats even when undervolted :()

MOBO            : MSI B350m Bazooka

MEMORY          : Team Group Elite TUF DDR4 3600 Mhz CL 16
STORAGE         : Seagate Baracudda 1TB and Kingston SSD
PSU             : Thermaltake Lite power 550W (Gonna change soon as i dont trust this)
CASE            : Rakk Anyag Frost
CPU COOLER      : ID-Cooling SE 207
CASE FANS       : Mix of ID cooling fans, Corsair fans and Rakk Ounos (planned change to ID Cooling)
DISPLAY         : SpectrePro XTNS24 144hz Curved VA panel
MOUSE           : Logitech G603 Lightspeed
KEYBOARD        : Rakk Lam Ang

HEADSET         : Plantronics RIG 500HD

Kingston Hyper X Stinger

 

and a whole lot of LED everywhere(behind the monitor, behind the desk, behind the shelf of the PC mount and inside the case)

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

yes, of course, its not my first rodeo :) the AIO cooler came with some pre applied paste wich i also cleaned and replaced with some mx4 arctic instead. pressure and ammount of paste are subjective, did the pea appling method but cant recall how big or how small it was, as for pressure i hand tighten the screws in a cross pattern until i felt it wouldn't budge. 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
1 minute ago, st4rway said:

yes, of course, its not my first rodeo :) the AIO cooler came with some pre applied paste wich i also cleaned and replaced with some mx4 arctic instead. pressure and ammount of paste are subjective, did the pea appling method but cant recall how big or how small it was, as for pressure i hand tighten the screws in a cross pattern until i felt it wouldn't budge. 

Not your first rodeo?? 

I've never experienced an AIO rodeo before. Didn't know you have either. Just trying to help.

 

Do you experience warm to hot air coming from the radiator or what? Could be just a matter of getting more intake. The bottom of that case with the fans restricts tons of air flow.

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

Actually my manual says otherwise, it depends on the model i think, it's plugged on my cpu fan header

mobo.png

TT manual.png

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author
20 minutes ago, ShrimpBrime said:

Not your first rodeo?? 

I've never experienced an AIO rodeo before. Didn't know you have either. Just trying to help.

 

Do you experience warm to hot air coming from the radiator or what? Could be just a matter of getting more intake. The bottom of that case with the fans restricts tons of air flow.

at the moment, only the lower 3 fans are providing intake, they are offset from the table only 1cm (0.39 inch) that's due to the vertical gpu mount limitation that don't allow me to mount them inside the case; the frontal 3 fans where the radiator is placed  and the top 3 fans are exaust. if i was to invert the orientation of the air flow on the radiator would that help? (considering it would actually put warmer air inside the case, but cooler air on the radiator)

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Dude....

 

The SYS FAN (PUMP) is right next to your 24 pin motherboard power plug.

 

 

 

121632304_TT360pumptosysfan.thumb.png.3f5100ca703586eac805bba78f79af93.png

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

that's not my aio cooler, that is an older model, mine only has one cable and according to the manual it is connected to the right header (i've included the correct page above) 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Well my 2 suggestions are as follows.

 

Better Cold air intake system (more fans, or remove restrictions) (I think maybe two fans back panel for intake, you have open space)

and

Plug the pump into the motherboard (pump) SYS FAN header.

 

that's all I got for you on this one. I think the rest of the setup is fine actually. Looks nice. Very clean. You did a fine job.

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Oh, let me explain about the difference between the Pump and cpu fan header.,

 

Firstly, the Cpu by default runs PWM and is controlled Via the Cpu temp. The Pump header has a higher current out put and is not generally defaulted to PWM.

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

So here's what in check for trying to solve the issue

- change the headers

- replace the termal paste

- change the rad airflow orientation

- change the GPU to horizontal and add some proper fans down there 

 

the back of the case unfortunately does not have fan mount holes and even if it did it only would allow 80 mm fans :s

 

thanks for help, will let you know if any of the sugestions worked

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
2 hours ago, st4rway said:

hi!

 

Been using this PC mostly for work and, after running into some PC freezes while playing HL Alyx i decided to find the root of the issue and did an prime95 cpu stress test.

while having the cpu at stock, with no OC whatsoever, i've reached package temps of 100°C wich are way too high for stock frequency.

Other than a possible not so good airflow issue i have absolute no idea whats going on and what to do to solve it.

 

Download HWinfo. and check your Vcore (don't be mistaken by VID, don't bother about it ) because all mobo at stock put really lots of voltage into CPU and that means lots of heat. so you could do negativ offset -0,040 or maybe more it depend on CPU ( everyone is diffrent) and LLC

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
45 minutes ago, st4rway said:

So here's what in check for trying to solve the issue

- change the headers

- replace the termal paste

- change the rad airflow orientation

- change the GPU to horizontal and add some proper fans down there 

 

the back of the case unfortunately does not have fan mount holes and even if it did it only would allow 80 mm fans :s

 

thanks for help, will let you know if any of the sugestions worked

I actually have a radiator that would fit that 80mm space I had used on my TJ07 case. Drill 4 mounting holes and a 5 $ fan. Quick easy intake fix. ;)

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author
4 hours ago, KotOr said:

 

Download HWinfo. and check your Vcore (don't be mistaken by VID, don't bother about it ) because all mobo at stock put really lots of voltage into CPU and that means lots of heat. so you could do negativ offset -0,040 or maybe more it depend on CPU ( everyone is diffrent) and LLC

Well, vcore was set to auto. If im not mistaken 9900k works somewhere beetween 1.3 and 1.4, and some cases with an good chip even under 1.3 with 5ghz overclock.

So i went testing, with realbench, and did an quick 15 min stress test at 1.32v, got max temps of 92C with stock frequencies! Tried lower and got blue screens, did i just get the wrong end of the silicon lottery? Or is it something else? 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
5 hours ago, st4rway said:

Well, vcore was set to auto. If im not mistaken 9900k works somewhere beetween 1.3 and 1.4, and some cases with an good chip even under 1.3 with 5ghz overclock.

So i went testing, with realbench, and did an quick 15 min stress test at 1.32v, got max temps of 92C with stock frequencies! Tried lower and got blue screens, did i just get the wrong end of the silicon lottery? Or is it something else? 

do you have latest bios? according to gigabite site it fix Vcore (but personally i would still try set negative offset) That 1.3V you tried was set manual ?

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Stock?  What frequency is reported as stock?  Stock is supposed to be 4.7 ghz on all cores, with a 1 and 2 core boost to 5 ghz (1 and 2 core loads only).


Can you please download HWINFO64 and report the CPU frequency shown at load, as well as monitor the VR VOUT voltage, as this is accurate vcore within +/-2mv.

You absolutely should *NOT* BSOD at stock at <1.32v under any conditions.

 

The only way this would happen is if you were not using any loadline calibration at all.  That's why you should check your VR VOUT at full load.

What is your BIOS set voltage when you get this BSOD?  1.30v?

 

Try the following.

Core : x47

Cache: x44

Bios voltage set: 1.250v

Vcore loadline Calibration: Turbo

CPU VRM switching frequency: 300 khz (this is important)

VCCIO: 1.15v

VCCSA: 1.20v

 

Disable XMP first and then test.

if stable, enable XMP and test again.

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

4.7 indeed, not even going to 5.0 on the 1st two cores...here's a print screen of the HWinfo after an quick 15 min stress test.

i kind of an noob when it comes to overclocking/voltage tweeking i can't seem to find the the CPU cache multiplier, CPU VRM switching and VCCSA under the bios advanced settings, would it possible to guide me towards it?

 

stress_test.png

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Did you remount the cooler yet?  Kinda sounds like a bad or loose mount.  When you say hand tighten -- do you mean with a screwdriver or your fingers?

AMD Ryzen 5800X|  Fractal Design S36 360 AIO w/6 Corsair SP120L fans  |  Asus Crosshair VII WiFi X470  |  G.SKILL TridentZ 4400CL19 2x8GB @ 3800MHz 14-14-14-14-30  |  EVGA 3080 FTW3 Hybrid  |  Samsung 970 EVO M.2 NVMe 500GB - Boot Drive  |  Samsung 850 EVO SSD 1TB - Game Drive  |  Seagate 1TB HDD - Media Drive  |  EVGA 650 G3 PSU | Thermaltake Core P3 Case 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

well not yet, i don't have any termal paste available at the moment, but already ordered some, expect to remount it tomorrow. 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
38 minutes ago, st4rway said:

4.7 indeed, not even going to 5.0 on the 1st two cores...here's a print screen of the HWinfo after an quick 15 min stress test.

i kind of an noob when it comes to overclocking/voltage tweeking i can't seem to find the the CPU cache multiplier, CPU VRM switching and VCCSA under the bios advanced settings, would it possible to guide me towards it?

 

stress_test.png

This is NOT HWinfo!  This is HWmonitor.  HWmonitor is garbage.

VR VOUT is in HWinfo64.

 

Also Realbench 2.43 is old and useless.  Please use 2.56, which supports AVX instructions.

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

intalled both HWinfo and the newer realbench, ran an quick stress test and according both to softwares temps did not go over 75C.

now on HWinfo data, what data should i look for? 

hwinfo.png

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
1 hour ago, st4rway said:

intalled both HWinfo and the newer realbench, ran an quick stress test and according both to softwares temps did not go over 75C.

now on HWinfo data, what data should i look for? 

hwinfo.png

 

I'm liking what I see here.

Temps look normal now.

7C temp delta between cores, and 1.135v load voltage at maximum load (minimum voltage because voltage will drop as amps (Current IOUT) gets higher).

(VR VOUT is the accurate die-sense voltage).

 

How did you reduce your temps so much?  Did you repaste and remount without telling us?

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

Nope, actually only reseted the bios and tweaked voltage a little...

Will replace the thermal paste with some kryonaut either way, will post results later

Link to comment
Share on other sites
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share
Go to topic listing CPUs, Motherboards, and Memory

×