Question Needing help with i9 10900k cooling :(

Feb 7, 2021
4
1
15
0
Hey!
This is my first time wielding such a powerful CPU and I was wondering if my Corsair H150i pro xt is enough ?
I expected temps to be lower because I saw on many videos that the same cooler + cpu combo achieved 65-70 Celsius avg in 1-3hrs of Cod Warzone.

I played my first hour of Apex Legends on 1080p locked at 235fps and Its smooth and all but my temps are around 80 Celsius all the time :(

My specs:
CPU: i9 10900k 4.9Ghz stock (bios mode to normal cuz temps are high in my opinion)
GPU: EVGA RTX3080 FTW3 ultra
SSD: Intel p660 1TB
PSU: Corsair 850W modular
RAM: Hyper X 2x8GB 3200mHz XMP1
CPU COOLER: Corsair H150i PRO XT
CASE: Corsair 680x Crystal

Fan locations:
360Radiator with 3 fans on the outer side of it are in FRONT Intaking fresh air
2x 140mm fan on top as exhaust
1x 120mm as exhaust in the back
2x 120mm on the bottom as intake

I don't know what I'm doing wrong I even bought an Arctic Cooler paste and I carefully operated in the building process... (paste was applied in an X shape)

Thankyou for your help in advance I'll post HWinfo screenshot down below!

cpu1
cpu2
cpu3
gpu
motherboard
 

uWebb429

Great
May 22, 2020
85
27
90
8
@Nosphere - 80°C is a normal operating temperature for all Intel CPUs. Intel sets the thermal throttling temperature to 100°C to protect the CPU from damage. Anything under that is fine. The temperature you are at now is not going to hurt anything so stop worrying. The maximum temperature is listed in the specs.

https://ark.intel.com/content/www/us/en/ark/products/199332/intel-core-i9-10900k-processor-20m-cache-up-to-5-30-ghz.html

Here is an example of the kind of abuse Intel's 10th Gen 10 core CPUs can take.



All 10 cores are fully loaded, power consumption is off the charts and the CPU is pegged at the thermal throttling temperature. It kept grinding away with no complaints, as it should be.

If you want to lower your temps, go into the BIOS and undervolt your CPU by -50 mV (-0.050 V). That can help knock close to 10°C off of your full load temps.
 
Feb 7, 2021
4
1
15
0
@Nosphere - 80°C is a normal operating temperature for all Intel CPUs. Intel sets the thermal throttling temperature to 100°C to protect the CPU from damage. Anything under that is fine. The temperature you are at now is not going to hurt anything so stop worrying. The maximum temperature is listed in the specs.

https://ark.intel.com/content/www/us/en/ark/products/199332/intel-core-i9-10900k-processor-20m-cache-up-to-5-30-ghz.html

Here is an example of the kind of abuse Intel's 10th Gen 10 core CPUs can take.



All 10 cores are fully loaded, power consumption is off the charts and the CPU is pegged at the thermal throttling temperature. It kept grinding away with no complaints, as it should be.

If you want to lower your temps, go into the BIOS and undervolt your CPU by -50 mV (-0.050 V). That can help knock close to 10°C off of your full load temps.
Yeah okay Ill try to lower the core voltage I was just a bit worried thankyou for clearing some things!
Do you think I can oc to 5.2Ghz allcore with a safe temp limit? Im not that experienced in adding 300mhz extra. I owned amd ryzen 3rd gen before this and that was terrible to oc
 
Reactions: TravisPNW

uWebb429

Great
May 22, 2020
85
27
90
8
A 10850K is a 10900K reject. Basically it is the exact same CPU but the 10850K needs extra voltage to be stable. Extra voltage equals extra power consumption and heat.

With a 10900K, you should be able to run an all core overclock of 5.2 GHz. If I reduce the voltage a little more, I can finish the Cinebench R20 benchmark at 5.2 GHz but the temps are going to be right around maximum, 100°C. These CPUs love to overclock but they get HOT. Sometimes I use a 53 multiplier when 4 cores or less are active and let it automatically switch to 52 when more than 4 cores are active.

I use ThrottleStop to change turbo ratios and voltages while in Windows. I have SVID Support enabled in the BIOS to allow ThrottleStop to control the voltages. This makes it easy to undervolt the CPU and do some testing without having to reboot.

This thread will give you an idea of what is possible when you tune these CPUs. They are very flexible.

https://linustechtips.com/topic/1286838-can-i-de-tune-my-i9-10850k-temporarily-for-heat-purposes/

ThrottleStop 9.2.9.7
https://drive.google.com/file/d/1AR37qxXO4GJsiY5atvCWgT97Rdp50UMl/view?usp=sharing
 
Reactions: TravisPNW

TravisPNW

Upstanding
Aug 26, 2020
359
131
360
0
Yeah okay Ill try to lower the core voltage I was just a bit worried thankyou for clearing some things!
Do you think I can oc to 5.2Ghz allcore with a safe temp limit? Im not that experienced in adding 300mhz extra. I owned amd ryzen 3rd gen before this and that was terrible to oc
A 10850K is a 10900K reject. Basically it is the exact same CPU but the 10850K needs extra voltage to be stable. Extra voltage equals extra power consumption and heat.

With a 10900K, you should be able to run an all core overclock of 5.2 GHz.
Agreed.

OP, You should be able to get to 5.2 all core... I got there with mine @ 1.34v with stress test temps on all cores in the 80-85C range. I read from many sources that pushing OCs above 85C isn’t recommended so I called it good there.

I run the Z73 AIO 360mm.
 
Reactions: Nosphere
Feb 7, 2021
4
1
15
0
Agreed.

OP, You should be able to get to 5.2 all core... I got there with mine @ 1.34v with stress test temps on all cores in the 80-85C range. I read from many sources that pushing OCs above 85C isn’t recommended so I called it good there.

I run the Z73 AIO 360mm.
I’ll try my best today and then let yall know how it went!
what stress test do u recommend ? I used cinebench r20 so far
 

TravisPNW

Upstanding
Aug 26, 2020
359
131
360
0
I’ll try my best today and then let yall know how it went!
what stress test do u recommend ? I used cinebench r20 so far
I used Cinebench... Blender benchmark... 3DMark... Prime 95 (small FFT)... FurMark and IntelBurnTest.

I called it good when P95 ran for an hour without any crashes. This was a month ago and haven’t had any issues since.

I did set an AVX offset of -2 (5ghz) just to keep temps in check when encoding with Handbrake. The difference in encoding time was only a couple of minutes...
 

CompuTronix

Intel Master
Moderator
Inside Intel's Secret Overclocking Lab: From page 7, paragraphs 9 & 10 ...

"At home, the lab engineers consider a load temperature above 80C to be a red alert, meaning that's the no-fly zone, but temps that remain steady in the mid-70’s are considered safe. The team also strongly recommends using adaptive voltage targets for overclocking and leaving C-States enabled. Not to mention using AVX offsets to keep temperatures in check during AVX-heavy workloads.

As Ragland explained, the amount of time a processor stays in elevated temperature and voltage states has the biggest impact on lifespan. You can control the temperature of your chip with better cooling, which then increases lifespan (assuming the voltage is kept constant). Assuming voltage remains constant, each successive drop in temperature results in a non-linear increase in life expectancy, so the 'first drop' in temps from 90C to 80C yields a huge increase in chip longevity. In turn, colder chips run faster at lower voltages, so dropping the temperature significantly by using a beefier cooling solution also allows you to drop the voltage further, which then helps control the voltage axis."

CT :sol:
 
Reactions: Phaaze88

Phaaze88

Polypheme
Ambassador
Inside Intel's Secret Overclocking Lab: From page 7, paragraphs 9 & 10 ...

"At home, the lab engineers consider a load temperature above 80C to be a red alert, meaning that's the no-fly zone, but temps that remain steady in the mid-70’s are considered safe. The team also strongly recommends using adaptive voltage targets for overclocking and leaving C-States enabled. Not to mention using AVX offsets to keep temperatures in check during AVX-heavy workloads.

As Ragland explained, the amount of time a processor stays in elevated temperature and voltage states has the biggest impact on lifespan. You can control the temperature of your chip with better cooling, which then increases lifespan (assuming the voltage is kept constant). Assuming voltage remains constant, each successive drop in temperature results in a non-linear increase in life expectancy, so the 'first drop' in temps from 90C to 80C yields a huge increase in chip longevity. In turn, colder chips run faster at lower voltages, so dropping the temperature significantly by using a beefier cooling solution also allows you to drop the voltage further, which then helps control the voltage axis."

CT:sol:
Oh, bananas! So things have changed, huh?
Went from 85C to 80C.
Manual > Adaptive to Adaptive > Manual.
C-states off to C-states on.

Wow, I somehow missed that article...
Well, I've got some work to do.
 
Last edited:
Reactions: CompuTronix

uWebb429

Great
May 22, 2020
85
27
90
8
Intel has been setting the thermal throttling temperature to 100°C for the vast majority of Core i CPUs that they have shipped since 2008. If this was causing problems, Intel would have lowered this temperature years ago to prevent early CPU deaths and warranty returns.

Intel's 10th Gen CPUs run great at high temperatures. The majority of high performance gaming laptops run the CPU at over 90°C for hours on end without any high rates of failure. If CPUs were prematurely failing, Intel would lower the throttling temperature in a heart beat.

so the 'first drop' in temps from 90C to 80C yields a huge increase in chip longevity.
Where is the test data to support this opinion? There is none. My closets are full of CPUs that are now too slow to be of any practical value. Not a dead one in the bunch. CPU failure due to heat exhaustion is a rare event. If this was happening on a regular basis there would be a lot more posts in the forums about this issue.

Why not listen to what Intel says. The PROCHOT (processor hot) signal goes active when the CPU has reached its "maximum safe operating temperature". For the 10850K, PROCHOT is set to a core temperature of 100°C.

 

CompuTronix

Intel Master
Moderator
CPU failure due to heat exhaustion is a rare event.
Your argument is, of course, valid, which is relative to page 7, paragraph 4 in the article I linked in my above post:

"premature chip death from overclocking isn't a common occurrence. "

so the 'first drop' in temps from 90C to 80C yields a huge increase in chip longevity.
Where is the test data to support this opinion?
The paragraphs I quoted in my first post originated from Dan Ragland, Principal Engineer, Performance Tuning & Overclocking Architecture at Intel's Jones Farm campus in Portland, Oregon, which, if you read the entire article I linked, explains their reasoning. If you require test data as proof, then you might consider contacting the author of our article, Paul Alcorn, whom may in turn, provide you with contact information for Dan Ragland, who may provide you with the test data in question.

Why not listen to what Intel says.
As we're both aware, Intel has published many "official" statements over the years which contradict with others, such as inconsistencies between the Data Sheets and the Product Specifications website. Here's a video of an interview between Tom's Hardware and Paul Zagacki, Principal Engineer, Client Computing Group from the 4790K erra, where 80°C is advocated, which is what was echoed by Intel's Engineers in our article.

Intel Discusses i7-4790K Core Temperatures and Overclocking

As the author of Real Temp and ThrottleStop, your contributions to the community, knowledge, thoughts and opinions have been widely respected and greatly appreciated throughout the years. However, as you've stated, as well, on other websites, anything below Throttle temperature is fine.

80°C is a normal operating temperature for all Intel CPUs.
While I agree that up to 80°C is fine, few are convinced or comfortable that operating their processors up to 99°C on a regular basis is sensible or acceptable.

The topic of Core temperature and voltage relative to processor longevity continues to be highly controversial, which has been hotly debated (pardon the pun) for decades, so we won't attempt to solve the debate in this Thread. Suffice it to say that in electronics, we can certainly all agree that there's nothing wrong about operating devices and their components at cooler temperatures, which experienced EE Engineers, as well as other academically related disciplines strongly advocate.

CT :sol:
 
Reactions: uWebb429
Feb 7, 2021
4
1
15
0
HEY ALL!!!
my particular problem was that my coolant temp went up fast cuz my cpu fan curves weren't adjusted correctly.
my cpu now runs on the same workload 20C cooler!!
However I only managed to push my cpu to a stable 5.0Ghz All Core 1.35V.
when I try to increase it to 5.2 or 5.3 my next stable Voltage is at 1.43 which is odd but it might come down to silicon lottery at this point right?
 

ASK THE COMMUNITY

TRENDING THREADS