[SOLVED] New i9-9900K 100 C Idle temperature, continued heating after restart

Jan 6, 2019
4
0
10
0
I found my new i9-9900k HOT, and I am both terrified and mad. Bought the CPU together with Gigabyte Z390 AORUS Elite before 3 days, and this is what I've experienced so far.

Now before I continue, I just want to let you know that the only thing I've done is clocking the RAM to 3200 because it was running on 2400. Nothing to worry about as I haven't touched anything else despite that. And, everything was running just fine. Also, I want to point out that I was HYPED and couldn't wait to get into many games I play to test the new CPU.

So I woke up in the morning, and there was plenty to find. I went to the toilet, which is close to the room where the PC is located and heard a loud noise produced by my coolers. I thought that NASA has come to my house and built in a rocket ready to launch to the moon. Seriously! I have 4 case fans including the 2 coming from the liquid cooler. You couldn't miss the loud noise at all. Liquid Cooling I use Cooler Master Master Liquid 240.

I launched HWmonitor and what to see... the view was quite unpleasant. The CPU was bolstering at 100 C, nothing more, nothing less. The thing that bothers me is that I do not know how long the CPU had been in that state. But, I have more to share on what I've done.

So, I instantly turned off the PC by unplugging the power cable, as any other normal human being would've done. After that, I waited for a couple of seconds and replugged it in. After starting the PC, went straight into the BIOS to check if everything was OK. 30 seconds after going into BIOS, the same noise started building and the temperature going towards the 100 C mark, increasing by 1 C each second. It was like a tempered bomb.

Getting the CPU temp that high after a power reset left me shocking. After acknowledging that, I loaded optimized defaults, saved and rebooted. That, helped me get the CPU temp back at a normal state. The temperature buildup was gone. That was the turning point towards good.

Unlike before, now I am using Z390 Gigabyte Aorus Elite. Not that I support or like them, but after being all my life with ASUS, I wanted to make a simple change and give some other company a chance to impress. Then I went to the official Gigabyte AORUS Elite support page and saw that there is a new BIOS to be downloaded. My retailer probably did not notice that there is a new BIOS available for that particular motherboard. Motherboard's current BIOS was 4 versions behind, and among the new bios releases, there was one updating the CPU microcode.

Without further thinking I downloaded and updated my BIOS to the latest version, something I should've checked and done before putting everything together. I always updated my BIOS before, and I do not know how I forgot to do the same in this case.

Therefore, I do not know who is up to blame, the Mobo or the Retailer. I assume this problem was caused by bad and inadequate BIOS. It could also be just a false temperature reading. But again, that is the case why I am making this thread.

Now, everything is running fine, temperatures are back at normal state.

Thing is, I am really scared as I do not know the exact timing of how long the CPU had been sitting at 100 degrees C without even shutting down by security measures. If anyone has any advice to add and help me ensure the stability of the CPU now that has been somewhat desecrated by unknown reason, please, let me know!

Also, would like this message to serve as a reminder to all the tech junkies who love to build their own PCs at home to never forget and check the BIOS Version.

P.S. Wife says that she did not hear any rumbling noise in the morning, so that's my only comfort zone and support to keep me happy enthusiastic.

Wish you a Happy New Year and thanks in advance for all your answers.
 

The Original Ralph

Distinguished
your Vcore voltage was probably way high,

As to any possible damage to your CPU, i doubt you've done it any harm. actually, intel's thermal limiter built into the cpu works pretty well at protecting the CPU. When i built my 4790(non-k) rig, i was relying on ASUS's AI Suite III perfromance utility for temp monitoring.

I used the rig for video rendering and noticed at 98-100% load, it seemed to always plateau out at 67C. The two rendering programs i use seem to be core hogs and run the CPU full tilt.

One day i downloaded RealTemp and then intel's XTU and both showed temps were actually running 98-100C, ie it would hit 100C, fall back to 98C then go right back up to 100C and continually repeat the cycle. I corrected the issue, ie better cooler, and started using the benchmark test in XTU to test my CPU, and as that test offers the opportunity to upload your benchmark score to the internet and compare it to other users with same CPU and mobo. Intel's thermal limiter will drop the CPU's load when temps hit self destruct level, that was why it would fall back to 97 or 98C, then climb back to 100C

- my scores are consistently in the top 5% - bear in mind, i rendered 1-2 video files a day, 2-3 hour jobs, 5 days a week, for over 3 months running it at 98-100C - that was 3 years ago. I'm pretty confident no permanent damage was done to the CPU

and a tip, before updating your BIOS again, do some research on the web and see if folks are experiencing any problems with the "latest / greatest" BIOS release before flashing it. And when you do flash, use the USB flash drive method, not the internet auto download or any other method.

Also, i'm not how you expect your retailer to flash the latest BIOS before selling you the board - you realize he'd have to install it in a computer to do that, and how many boards does he have the time to do that to?

something else that occurs to me, i went thru two Z390 GIgabyte motherboards - both, the BIOS was self-corrupting and change values on their own. MSI has a pretty decent "OC by the numbers" guide on their website https://www.msi.com/blog/intel-9th-cpu-overclocking-5ghz-with-z390-motherboards - Use it to learn your BIOS and everytime you turn that computer on, check your Vcore, make sure it hasn't changed. I'd do that until i was comfortable the BIOS is stable

while i doubt hitting those temps hurt your CPU, it's not really recommended to test it like that
 

The Original Ralph

Distinguished
your Vcore voltage was probably way high,

As to any possible damage to your CPU, i doubt you've done it any harm. actually, intel's thermal limiter built into the cpu works pretty well at protecting the CPU. When i built my 4790(non-k) rig, i was relying on ASUS's AI Suite III perfromance utility for temp monitoring.

I used the rig for video rendering and noticed at 98-100% load, it seemed to always plateau out at 67C. The two rendering programs i use seem to be core hogs and run the CPU full tilt.

One day i downloaded RealTemp and then intel's XTU and both showed temps were actually running 98-100C, ie it would hit 100C, fall back to 98C then go right back up to 100C and continually repeat the cycle. I corrected the issue, ie better cooler, and started using the benchmark test in XTU to test my CPU, and as that test offers the opportunity to upload your benchmark score to the internet and compare it to other users with same CPU and mobo. Intel's thermal limiter will drop the CPU's load when temps hit self destruct level, that was why it would fall back to 97 or 98C, then climb back to 100C

- my scores are consistently in the top 5% - bear in mind, i rendered 1-2 video files a day, 2-3 hour jobs, 5 days a week, for over 3 months running it at 98-100C - that was 3 years ago. I'm pretty confident no permanent damage was done to the CPU

and a tip, before updating your BIOS again, do some research on the web and see if folks are experiencing any problems with the "latest / greatest" BIOS release before flashing it. And when you do flash, use the USB flash drive method, not the internet auto download or any other method.

Also, i'm not how you expect your retailer to flash the latest BIOS before selling you the board - you realize he'd have to install it in a computer to do that, and how many boards does he have the time to do that to?

something else that occurs to me, i went thru two Z390 GIgabyte motherboards - both, the BIOS was self-corrupting and change values on their own. MSI has a pretty decent "OC by the numbers" guide on their website https://www.msi.com/blog/intel-9th-cpu-overclocking-5ghz-with-z390-motherboards - Use it to learn your BIOS and everytime you turn that computer on, check your Vcore, make sure it hasn't changed. I'd do that until i was comfortable the BIOS is stable

while i doubt hitting those temps hurt your CPU, it's not really recommended to test it like that
 
Jan 6, 2019
4
0
10
0


Ofc I used the usb flash drive method. But, the problem is that the PC was already turned on when it accidentally entered this state. I just woke up to find it like that, and turned it off instantly -> reset bios -> update bios.

Hopefully, this won't be an issue anymore. About the CPU, ofc I am feeling bad because it's just 3 days old. It is strange to see the vCore going sky high out of nowhere when before going to bed I left it in a normal state.+

Anyhow, thanks for your answer.
 

CompuTronix

Intel Master
Moderator

angerzr,

On behalf ot Tom's Moderator Team, welcome aboard!

• You have an intermittent flow problem with your AIO.

The BIOS issue is merely coincidental and unrelated, as the problem corrected itself before you updated. Moreover, although Ralph's theory concerning "runaway" Vcore due to BIOS corruption is interesting, only once have I ever read of such an instance, which resulted in BSOD's and rapid CPU failure.

Concerning your AIO, it happened "during the night ... or ... I woke up to loud fans ... or ... out of nowhere ... or ... it was fine last night" is a common scenario which is very typical of a flow blockage or faulty pump, not high Vcore. Core temperatures just don't jump to throttle temperature at 100°C overnight unless there's an AIO problem, which often occurs suddenly for no apparent reason.

There are four problems with AIO's:

(1) All AIO's will eventually fail. It’s not a question of if; it’s a question of when. Pumps can have component failures in the electronics that drive the motor. Pumps also have moving parts (motor and impeller) that wear out, so those which run 24/7/365 are prone to premature failure. AIO units are notorious for failures due to inferior pump quality, whereas custom loops typically use high-end pumps which have greater longevity.

(2) Coolant can evaporate over time due to "permeation" which can cause impeller "cavitation" due to bubbles.

(3) AIO's are sealed Closed Loop Coolers (CLC), which unlike custom loops, are not designed to be disassembled, the water block and impeller flushed and cleaned of bio-contaminants and radiator sediments, then refilled.

(4) Also unlike custom loops, AIO's use dissimilar metals (aluminum radiator / copper water block). This causes galvanic corrosion which produces sediments that accumulates over time, resulting in blockages and flow problems. Even new AIO's may contain radiator sediments due to inadequate flushing after manufacturing. Experienced builders of custom loops always thoroughly flush brand-new radiators. Flux, solder and metal fragments are typically found when the flush water is poured through a strainer.

For no apparent reason ... OR ... by simply installing, moving, tipping, handling or otherwise disturbing a NEW OR OLD unit, particles can become dislodged, whereupon the next power-up, the now free-floating particles can circulate into the impeller chamber and restrict or completely stop the impeller. Since the impeller is magnetically coupled to the motor (no direct shaft), the motor will continue to run while the impeller is stalled or stopped. So "hearing" the motor running or "feeling" vibration does not indicate flow, nor does Pump RPM in various software utilities.

Without an actual in-line sensor, proper flow is difficult to verify. Under normal operation, even at 100% workload, there should be only a minimal temperature differential between the tubes. However, if one tube is hot while the other is cool, or the water block is hot while the radiator is cool, it indicates little to no flow. This can be verified by checking various surface temperatures with an infrared (IR) thermometer.

• It's highly likely that your AIO contains radiator sediments from manufacturing, which caused an intermittent impeller stall. Since the problem occurred once, you can expect that it will happen again.

As your AIO is probably new, it should still in warranty. I suggest that until you get a replacement cooler, that you don't leave your rig powered up or unattended when not in use.

Once again, welcome aboard!

CT :sol:
 

The Original Ralph

Distinguished
That's probably the best dissection of water cooling (especially the pump part) that i've seen. I wasn't aware the pumps were magnetically driven, and definitely good to know. You've convinced me to add at least a flow indicator if not a flow meter to my system i'm assembling. But your deductions sure appear solid

but i will add one comment, over volting from a corrupted BIOS was what fried my Asrock Taichi Z390 and my CPU - both the Asrock tech and the tech at Overclockers, indicated that after testing my CPU - only thing he could figure was over voltage. And btw i believe the OP indicated temps normalized after he loaded default settings & rebooted, not when he updated his BIOS, leaves wonder ing if BIOS didn't play a part, either with voltage or his AIO pump
 

CompuTronix

Intel Master
Moderator
Ralph,

Thank you for that information. Now that you mention it, I recall reading another of your posts where you shared that unfortunate incident which occurred with your shiny new 9900K.. :( Now I'm aware of two such Vcore problems.

The instance to which I referred was interesting in that BIOS overvolted the CPU by exactly 200 millivolts, so the voltage set in BIOS was not the actual voltage applied to the cores. The processor degraded within minutes, followed by BSOD's after every boot. It was a QX9650 (quite pricey back in the day) on an ASUS P35 Deluxe motherboard. That 45nm processor was subjected to 1.6 Vcore when they should never receive more than 1.4. :ouch:

The problem followed an indirect nearby lightning strike. It didn't interrupt power, yet it's interesting that the computer was connected to a UPS, but wasn't turned on. Regardless, there's no doubt that the lightning strike corrupted BIOS, as upon the next powerup, it was the beginning of a quick end.

However, in both your instance and that which I just described, immediate CPU failure was the result. But in angerzr's case, the circumstances are different, and instead conform to the same scenario we see in our forums all too often; "I woke up to loud fans and 100°C". I think at the present state of AIO reliability, the problem which angerzr characterized speaks clearly.


With respect to the chicken or the egg, I believe I correctly interpreted the order of events which I condensed as "turned off ... replugged it in ... temperature increasing by 1 C each second ... loaded optimized default ... rebooted ... temp back at normal ... updated BIOS".

I took that to mean;

100°C found > cycled power > 1°C ^ / sec > (cycled power) loaded default > reboot (cycled power) > temp normal > updated BIOS.

Temp normal precedes updated BIOS.

Regardless, the significance is that power was cycled possibly three times. The first power cycling didn't affect the problem, which was cleared after the second power cycling. When the power is cycled, the pump cycles with it. If the impeller is stalled by an obstruction between a blade and the chamber, it can be freed when the pump is stopped and restarted. The result is a puzzling "intermittent" (12 letter word with a 4 letter meaning) temperature problem. We see many such threads.


Concerning the potential for a BIOS related Vcore problem, since angerzr's Core temperatures returned to normal ... for the present ... and his CPU survived ... given the symptoms, I think there's enough evidence to rule out BIOS and Vcore issues.

This thread is yet another unfortunate example of typical AIO behavior.

Although I run an AIO in my personal rig for the purpose of comparing thermal performance, as well as for aesthetics, my big air cooler is always kept close at hand as a backup, which I think intimidates the AIO so it'll keep working. :D

CT :sol:
 

The Original Ralph

Distinguished
i don't know that i totally agree about voltage not being the culprit, but given you're much more literate in these matters than i am, i'll rest with your judgement

to the OP - suggestion, take that AIO back to the seller, and get an EKWB unit - it will be better quality plus have the ability to expand to cool other components. As it is, you can't take that one apart, at least not without voiding your warranty

Computronix, so you are aware, and i suspect you are, i've gone thru 3 Z390 mobos, 1` was the Asrock mentioned above, and the other 2 were Gigabyte Aorus Master boards, both of which demonstrated self corrupting BIOSs

but to your comment about leaving an air cooler on the shelf near the computr for your AIO to see it, i once brought a sledge hammer into the room during the days of Windows 95 - after 2 crashes inside of 30 days, each causing 11-12 hours rebuilding seesions, i told the computer "Meet Mr Sledge, Mr Sledge meet my computer" and to the computer i literallly said, if you don't pull your weight, you two will get intimate. But as someone said, there were 3 major catastrophes in the 20th Century, 1) Hiroshima, 2) Chernobyl and 3) Windows 95
 

jankerson

Judicious
BANNED


The MB makers have been completely idiotic with their BIOS for the 9th Gen CPU's, crazy insane high VCORE.

I had the same thing with my Z370 AORUS GAMING 5 with the supposed 9th Gen optimized F10 and F11 BIOS. CRAZY high VCORE over 1.4V STOCK and over 1.5V MC Enhancement... Stupid idiotic morons.... I went back to the F7 BIOS and everything was normal again.

And they know they have issues with the BIOS on the Z370 and Z390 MB's with 9th Gen CPU's with the crazy VCORE values.

The real clue was the VCORE went nuts with the 8086K when I updated it before hand. Changed it back to F7 and then F10 and F11 after I installed the 9900K. Same thing like I said above so I had to go back to F7, THAT WORKS GREAT>

I am just glad I have a rock solid MB with solid VRMS so I don't have to worry about it.



 

The Original Ralph

Distinguished
i've been noticing quite a few posts re BIOSs that were not acting consistently. I mentioned to someone else, on the Z390 BIOSs, a number of mfgrs seemed to, as the Brits say, 'missed the plot" on the Z390 BIOS coding.

a world class OCer on another forum, hellaciously nice guy & former Asus & MSI engineer, recommended the MSI board in my sig and i love it - i've actually left it at the bios release it came with -rock solid stable
 

jankerson

Judicious
BANNED


Maybe they will get their heads out of their *&^%$ and fix it some day across the board makers.

I talked to GB support and got the basic blah, blah, blah... BS Excuses....

Then talked to AORUS support and got the real story, they know about the issues and are supposed to be working on them.

 

The Original Ralph

Distinguished
on the Asrock Taichi Z390 that literally fried itself and my CPU, John the tech said 7-10 day turn around on the board - they rcvd it Nov 10th, and i can't even get a status report via the very email they sent me to use if had any questions. Got a feeling they're up to their necks in returned boards
 
Jan 6, 2019
4
0
10
0
I've read your opinions. I just want to point it out that I truly believe it had to be a corrupted BIOS. Because A: The AIO is not new and I am using it for 3-4 months and never experienced this before, B: Disconnecting the PSU by unplugging the power cable did not change the behavior. Instead, resetting the BIOS did.
After that, I just updated the BIOS to the latest version and that was about it. Now, everything runs fine for 24 hours straight. I am just scared about the longevity of the CPU. I am closely monitoring it and I am scared of experiencing any similar SURGE in the future.

Hopefully, it won't happen and it was caused by a bad BIOS that managed to mess up the VCORE. Hopefully...

Thanks for your replies by the way. Appreciate it a lot!
 

CompuTronix

Intel Master
Moderator
Your CPU is fine, so you needn't be concerned about it.

Intel's "Tj Max" thermal specification, which is also called "Throttle" temperature is 100°C. This is the Core temperature limit at which the processor will reduce Core voltage and Core speed to prevent thermal damage. It's an internal safeguard that Intel has engineered into their processors for more than a decade. It works, exactly as Ralph explained in his first post.

If your CPU had been seriously overvolted, it would not have survived.

Respectfully, believe what you prefer, or disregard what you don't want to hear, but either way, AIO's can become intermittent or outright fail at any time, new or otherwise. The age of the unit means nothing, except for RMA purposes. Every day we see the same scenario you described, repeatedly unfold on our forums. That's why Ralph runs the best air cooler money can buy.

My best advice to you is don't leave your rig powered up or unattended when not in use, and always monitor your Core temperatures.

I hope it works out for you.

CT :sol:
 

The Original Ralph

Distinguished
a suggestion then - download RealTempGT and install it. It's a temp monitoring utility. If you double clik on it up near the top it'll reduce to a widget sized window, but there's an alarm feature - i have mine set for 85C. THe alarm it sounds if your computer hits that temp for a couple of seconds will wake the dead, so at least if you're in the other room if it happens again, you'll know to get back to the computer
 

CompuTronix

Intel Master
Moderator

That alarm is an audio clip from the original Star Trek series from the `60's. It's the alarm for "Red Alert" Battle Stations. When Kevin Glynn, the author of Real Temp, was developing his software, he was looking for an appropriate alarm, so I uploaded "Red Alert" to him. He liked it, so that was my big contribution to Real Temp.

CT :sol:
 
Jan 6, 2019
4
0
10
0


Well, that would probably be the best option for now, until this whole mess gets resolved. Either way, I am really really happy and somewhat encouraged, pleased because of your answers. After all, it's not the end of the world. Thanks a lot, guys!
 

ASK THE COMMUNITY