Accidentally ran the 10900K with Asus Strix Z490-I at 100C

[Mountainlifter]

I was doing some cable management and disconnected the pump block's power and forgot to connect it again.

Fans spun really high when I booted the system and once I was in windows, I immediately loaded Hwinfo to find out why. The CPU temps were at 99/100C.

I immediately shut down my PC. Even the pump block was hot to the touch after this happened.

The Bios settings were on default with MCE on AUTO but I'd enabled XMP for the RAM. I haven't gotten around to OC'ing or anything else yet.

Did I cause any damage to my CPU, Mobo or Pump block? Should I be worried and can I run some tests?

In all, it ran this way for 4 to 5 minutes at most at 100C.

[Arne_Saknussemm]

The CPU will throttle if it is too hot. And you have a small amount of thermal capacity in the block and water will move by convection slightly....so....all in all you should be OK if it was a short excursion

Maybe remount the block and repaste....if you have the tools/inclination...

Tests? Just plug in the pump and see what's what temp wise at idle and at load....check spread of temps between cores

[Mountainlifter]

Arne Saknussemm wrote:
The CPU will throttle if it is too hot. And you have a small amount of thermal capacity in the block and water will move by convection slightly....so....all in all you should be OK if it was a short excursion

Maybe remount the block and repaste....if you have the tools/inclination...

Tests? Just plug in the pump and see what's what temp wise at idle and at load....check spread of temps between cores

I did a repasting but the previous pasting was good too I noticed when I opened it up.

The temps are fine at idle and games send it to 60 to 70C.

A couple of cores seem hotter than the others at idle by 7 to 10C on average but I'm guessing they're getting assigned more tasks.

Looks like the main thing hurt was my pride at making this dumb mistake. I shouldn't have meddled with it when i was tired and it was late. 😞

I would say it was a short excursion to 100C and I didn't run any stressful load. Only viewed the bios for a couple of minutes and booted into the OS for another couple of minutes. Anyways are there some tests I can run to check the cpu's health or the motherboard's health?

[eegee]

My core temps vary around 5-10°. HW-Info and CPU-Z.
For 10900K it is normal, that only 2-4 cores are on Extra-Turbo (x53) mode. Normal is max x50, by cooling one core on x8.

There is the “Intel Extreme Tuning Utility” on ASUS drivers page (Maximus XII board). If you find it for your board, you can test your machine under stress with that.
Also, there is a “Intel Processor Diagnostic Tool” on Intels Webpage. Just for the nerves to calm down.

Many of us were there, in one way or the other.

[Mountainlifter]

eegee wrote:
My core temps vary around 5-10°. HW-Info and CPU-Z.
For 10900K it is normal, that only 2-4 cores are on Extra-Turbo (x53) mode. Normal is max x50, by cooling one core on x8.

There is the “Intel Extreme Tuning Utility” on ASUS drivers page (Maximus XII board). If you find it for your board, you can test your machine under stress with that.
Also, there is a “Intel Processor Diagnostic Tool” on Intels Webpage. Just for the nerves to calm down.

Many of us were there, in one way or the other.

Thanks for the reply and suggestions.

Just did a thorough repasting and booted up the system. Intel stock settings with MCE disabled and XMP I enabled. These are my idle temps after 30 minutes. https://imgur.com/otRtsme

These are the temps while running the “Intel Processor Diagnostic Tool” https://imgur.com/j5Dzm9T The tool says all tests Pass. I will try the Intel Extreme tuning utility soon.

Cores 4 and 5 seem to have the maximum usage which is why their temps are higher.

So, temps seem fine even if they are a little high because I am using a 240mm Kraken X53 cooler with an SFF case (Ncase M1). And I live in Singapore where the midday ambient temp. is 32C and relative humidity is between 80 to 90% .

Guess I just have to look at the bright side on this one. It didn't run at 100C for more than 5 minutes. there was no auto-shutdown activated by the CPU. There was no smell of burning money. And I learnt a valuable lesson.

[Falkentyne]

Mountainlifter wrote:
Thanks for the reply and suggestions.

Just did a thorough repasting and booted up the system. Intel stock settings with MCE disabled and XMP I enabled. These are my idle temps after 30 minutes. https://imgur.com/otRtsme

These are the temps while running the “Intel Processor Diagnostic Tool” https://imgur.com/j5Dzm9T The tool says all tests Pass. I will try the Intel Extreme tuning utility soon.

Cores 4 and 5 seem to have the maximum usage which is why their temps are higher.

So, temps seem fine even if they are a little high because I am using a 240mm Kraken X53 cooler with an SFF case (Ncase M1). And I live in Singapore where the midday ambient temp. is 32C and relative humidity is between 80 to 90% .

Guess I just have to look at the bright side on this one. It didn't run at 100C for more than 5 minutes. there was no auto-shutdown activated by the CPU. There was no smell of burning money. And I learnt a valuable lesson.

Core #2, 4 and 6 are always the hottest cores on every 10900k CPU (also applies to 9900k). Core #1 is always the coolest core in prime95 SSE2/AVX/FMA3 small FFT. For some reason, this changes a bit in LinX/Linpack/Aida64 "Stress FPU" with a couple of the cores. But in heavy sustained loads, its core 2,4,6 hottest, followed by core 0, and cores 7,8,9 are usaully about 1C from each other and core 1 the absolute coolest. Has to do with the convexity of the die layout combined with something with the chip/cache layout which is beyond me.

[Mountainlifter]

Falkentyne wrote:
Core #2, 4 and 6 are always the hottest cores on every 10900k CPU (also applies to 9900k). Core #1 is always the coolest core in prime95 SSE2/AVX/FMA3 small FFT. For some reason, this changes a bit in LinX/Linpack/Aida64 "Stress FPU" with a couple of the cores. But in heavy sustained loads, its core 2,4,6 hottest, followed by core 0, and cores 7,8,9 are usaully about 1C from each other and core 1 the absolute coolest. Has to do with the convexity of the die layout combined with something with the chip/cache layout which is beyond me.

Thanks for the info. I will lookout for these trends.