Overclock.net banner

[6600K] I don't recall my CPU being this hot...is this normal?

2K views 48 replies 6 participants last post by  GridIroN 
#1 ·
I was testing my 6600K overclock after updating the BIOS on my STRIX when I realized my temperatures seemed high. After doing some testing, I'm like some feedback as it seems to my like my temps are too high and something's wrong...

Static environment:
Asus Z270 Strix [BIOS 1009]
Corsair H100i V1 AIO cooler [EK Vardar fans @100%]
Ambient temperature [22 C]

Software
Cinebench R11.5
Prime 95 v29.2
HWinfo v5.56

STOCK
speed [3.5ghz]
Vcore [1.250v]
XMP [on]
VCCIO [1.125v]
VSSA [1.125v]

VID [1.210v]

Cincebench [6.3 pts @ 63c]
Prime95 [77c]
Super Position [40% load]-[45c]

OVERCLOCKED
speed [4.4ghz]
Vcore [1.350v]
XMP [on]
VCCIO [1.125v]
VSSA [1.125v]

VID [1.365v]

Cincebench [7.78 pts @ 85c]
Prime95 [88@c]
Super Position [40% load]-[58c]

Temps seem within normal range, though high in normal gaming situations, like Superpositions, but if I give the chip any serious load, the temps seem to skyrocket to TjMax....I've had the chip since the 6 series released, so it's not new.
 
See less See more
#2 ·
Overtime things such as bolts and screws might get lose or more relaxed, I've experienced something similar with some of the rigs I've worked with. Nothing to be concerned about, simply reseat + reapply new paste should fix your "issue".
 
#3 ·
Also dust can get into the system and you may find it is starting to block the inlet face of the h100 - nothing a good vacuum won't fix.
 
#5 ·
Are the voltages listed their set points in bios, or the readings in os? What program are you using to monitor? Have you updated it recently as well?

Have you checked PLL voltage?

When was the last time you checked temps before the update? When was the last time you cleaned the radiator and fans, filters? On my HTPC that only has one 120 intake, the filter needs cleaning at least every two weeks or water temp starts rising on the aio.
 
#6 ·
Quote:
Originally Posted by Boyd View Post

Overtime things such as bolts and screws might get lose or more relaxed, I've experienced something similar with some of the rigs I've worked with. Nothing to be concerned about, simply reseat + reapply new paste should fix your "issue".
Perhaps I'll try a full reseat soon and see if anything was loose.
Quote:
Originally Posted by steveting99 View Post

What is your idle numbers? Please post a screen shot using HWiFO64.

Please run Asus Realbench stability test with HWiFO64 and report back what the numbers are. Screen shot would be useful.

Thanks.

Quote:
Originally Posted by EMUracing View Post

Are the voltages listed their set points in bios, or the readings in os? What program are you using to monitor? Have you updated it recently as well?

Have you checked PLL voltage?

When was the last time you checked temps before the update? When was the last time you cleaned the radiator and fans, filters? On my HTPC that only has one 120 intake, the filter needs cleaning at least every two weeks or water temp starts rising on the aio.
As main post says, HWinfo, CPU-Z, MSI Afterburner, RealTemp etc. I get lower idles on HWinfo, but generally they all report the same loads.

I had to take my PC apart when I got my new motherboad, I'd say earlier this year, before summer, so probably 3-4 months.
 
#7 ·
Quote:
Originally Posted by GridIroN View Post

As main post says, HWinfo, CPU-Z, MSI Afterburner, RealTemp etc. I get lower idles on HWinfo, but generally they all report the same loads.

I had to take my PC apart when I got my new motherboad, I'd say earlier this year, before summer, so probably 3-4 months.
With the previous BIOS, do you have a screen shot of HWiFO64 idle numbers so one can compare? The idle numbers for overclock seems a bit high, with individual cores running in the high 40's and CPU package over 50 deg C. Is this after 5 minutes when the system has settled down? If the ambient is 22 deg C, the delta for the CPU cores is around 25 deg C and the CPU package is close to 30 deg C, and this is at idle condition! I would be looking at the surface contact between the CPU IHS and the cold plate of the Corsair H100i V1 AIO if it is firm.

Core usage is in the high single digits and low teens under idle condition?

At idle conditions, why are the fans operating at high RPM? So is the CPU / AIO? If the fans on the radiator for the Corsair H100i V1 AIO are near 100%, there doesn't appear to be much thermal headroom available once the cores start to fully load. Conversely, if the AIO radiator fans and pump are running at 100%, the liquid cooling loop should be a close approximation of the internal case temp during idle condition. Does Corsair provide a water thermal sensor on their AIO which HWiNFO64 can extract? It would be useful.

What is the internal case temp? With room ambient at 22 deg C, one of the HDDs have heated up to 45 deg C? What is the delta between internal case temp and room ambient?

Your RAM usage appears high. Enough so that Windows is using the page file on your system drive. With all your normal programs running after startup, what is the % free RAM available?

Edit: Item in bold refers. Are you comparing differences between motherboards?
 
#8 ·
Quote:
Originally Posted by steveting99 View Post

With the previous BIOS, do you have a screen shot of HWiFO64 idle numbers so one can compare? The idle numbers for overclock seems a bit high, with individual cores running in the high 40's and CPU package over 50 deg C. Is this after 5 minutes when the system has settled down? If the ambient is 22 deg C, the delta for the CPU cores is around 25 deg C and the CPU package is close to 30 deg C, and this is at idle condition! I would be looking at the surface contact between the CPU IHS and the cold plate of the Corsair H100i V1 AIO if it is firm.
I don't have any screenshots of HWinfo from the previous BIOS. My temperatures were normal though. 40ish at idle with manual voltage, never downclocking and 65'ish on load.
Quote:
Originally Posted by steveting99 View Post

At idle conditions, why are the fans operating at high RPM? So is the CPU / AIO? If the fans on the radiator for the Corsair H100i V1 AIO are near 100%, there doesn't appear to be much thermal headroom available once the cores start to fully load. Conversely, if the AIO radiator fans and pump are running at 100%, the liquid cooling loop should be a close approximation of the internal case temp during idle condition. Does Corsair provide a water thermal sensor on their AIO which HWiNFO64 can extract? It would be useful.
I set them 100% to test if it was the AIO. If the fans are at 100% it wouldn't require any more headroom as the cooler is operating at 100% of it's ability to dissipate heat.
Quote:
Originally Posted by steveting99 View Post

What is the internal case temp? With room ambient at 22 deg C, one of the HDDs have heated up to 45 deg C? What is the delta between internal case temp and room ambient?
It's a Phantek EVOLV ATX, it's not a cold case to begin with, but my maintenance archive of apps is on a mechanical HDD. It was running the tests so it heated up most likely.
Quote:
Originally Posted by steveting99 View Post

Your RAM usage appears high. Enough so that Windows is using the page file on your system drive. With all your normal programs running after startup, what is the % free RAM available?
Not sure. Ill have to check and report back. I usually have 11GB free of my 16.

Edit: Item in bold refers. Are you comparing differences between motherboards?[/quote]
 
#9 ·
Quote:
Originally Posted by GridIroN View Post

I don't have any screenshots of HWinfo from the previous BIOS. My temperatures were normal though. 40ish at idle with manual voltage, never downclocking and 65'ish on load.
I set them 100% to test if it was the AIO. If the fans are at 100% it wouldn't require any more headroom as the cooler is operating at 100% of it's ability to dissipate heat.

It's a Phantek EVOLV ATX, it's not a cold case to begin with, but my maintenance archive of apps is on a mechanical HDD. It was running the tests so it heated up most likely.

Not sure. Ill have to check and report back. I usually have 11GB free of my 16.

Edit: Item in bold refers. Are you comparing differences between motherboards?
Can you take a pic of the inside of your case and post? It would be good to see how you've set up the AIO, fans and look at the airflow. I.e. where is the AIO in the case and is it configured as intake or exhaust?

Asus Strix Z270E Gaming has a connection for a thermal sensor called T_Sensor1. You can grab a cheap thermal sensor from Amazon such as the XSPC 10k thermal sensor to measure the internal (or device) temperature. If you don't know what the internal temp of the case is, this will allow you to monitor and record using HWiNFO64. With Asus Fan Expert, you can tie in T_Sensor1 to startup/shutdown of the fans - a useful feature.

If things were okay with the previous BIOS version, Asus has a handy USB flash utility that will allow you to flash the board to the previous BIOS. Just write down and capture these results for the current BIOS settings. Carry out Asus Realbench stability tests for default and overclocked conditions. Use HWiNFO64 to capture the results.

Once you have flashed the previous BIOS, you can easily do the idle and stability tests of the system under default and overclock conditions. Then compare the results if the issue is really the BIOS or something else.

You mentioned that 3 to 4 months ago, the motherboard had changed? Doing so would most likely involve re-seating the CPU. Below are some pictures of a good fit between CPU IHS and the cold plate of your AIO.


When applying the TIM, keep in mind that less is better. The reason being is the thermal paste is meant to fill the void between the metal to metal contact between the CPU IHS and cold plate of your AIO as shown below.
 
#10 ·
Quote:
Originally Posted by steveting99 View Post

Can you take a pic of the inside of your case and post? It would be good to see how you've set up the AIO, fans and look at the airflow. I.e. where is the AIO in the case and is it configured as intake or exhaust?

Asus Strix Z270E Gaming has a connection for a thermal sensor called T_Sensor1. You can grab a cheap thermal sensor from Amazon such as the XSPC 10k thermal sensor to measure the internal (or device) temperature. If you don't know what the internal temp of the case is, this will allow you to monitor and record using HWiNFO64. With Asus Fan Expert, you can tie in T_Sensor1 to startup/shutdown of the fans - a useful feature.

If things were okay with the previous BIOS version, Asus has a handy USB flash utility that will allow you to flash the board to the previous BIOS. Just write down and capture these results for the current BIOS settings. Carry out Asus Realbench stability tests for default and overclocked conditions. Use HWiNFO64 to capture the results.

Once you have flashed the previous BIOS, you can easily do the idle and stability tests of the system under default and overclock conditions. Then compare the results if the issue is really the BIOS or something else.

You mentioned that 3 to 4 months ago, the motherboard had changed? Doing so would most likely involve re-seating the CPU. Below are some pictures of a good fit between CPU IHS and the cold plate of your AIO.


When applying the TIM, keep in mind that less is better. The reason being is the thermal paste is meant to fill the void between the metal to metal contact between the CPU IHS and cold plate of your AIO as shown below.
I appreciate your time and resources, but perhaps it would save time and effort for me to iterate that I'm not a noobie. I've been building rigs since I was 14 (now 29) and as my sign suggests, I've even won Guru3D ROTM. I'm 143% familiar with the concept of concave metallurgy at atomic levels. I'm posting here about this because I'm completely baffled as to what could have caused this.

I can inform you that I just finished a full reseat of my H100i and my temperatures haven't changed one bit. Also, all my cores are uniformally hot, so I can rule out the H100i itself, or a bad TIM seating. The issue must be with a BIOS setting, or the chip itself...

I may have to spend tomorrow reflashing BIOS v906 and see if it's the BIOS to rule that out as well.
 
#11 ·
That's a very high vcore for 4.4ghz

still very hot, around 20c hotter than my 6700k running 4c4t on air cooling at the same volts. What's your ambient temp around? Try setting XMP off, sometimes weird stuff happens there and you have it on the "stock" profile even though XMP is overclocking and overvolting some things
 
#12 ·
Quote:
Originally Posted by Cyro999 View Post

That's a very high vcore for 4.4ghz

still very hot, around 20c hotter than my 6700k running 4c4t on air cooling at the same volts. What's your ambient temp around? Try setting XMP off, sometimes weird stuff happens there and you have it on the "stock" profile even though XMP is overclocking and overvolting some things
22 C
 
#13 ·
Quote:
Originally Posted by GridIroN View Post

I appreciate your time and resources, but perhaps it would save time and effort for me to iterate that I'm not a noobie. I've been building rigs since I was 14 (now 29) and as my sign suggests, I've even won Guru3D ROTM. I'm 143% familiar with the concept of concave metallurgy at atomic levels. I'm posting here about this because I'm completely baffled as to what could have caused this.

I can inform you that I just finished a full reseat of my H100i and my temperatures haven't changed one bit. Also, all my cores are uniformally hot, so I can rule out the H100i itself, or a bad TIM seating. The issue must be with a BIOS setting, or the chip itself...

I may have to spend tomorrow reflashing BIOS v906 and see if it's the BIOS to rule that out as well.
As you've got test results for idle and load temps on the current BIOS, it would not be that difficult to flash the older BIOS and re-test.

If the test results are the same, check the coolant level on the H100i.

[Edit: See the following thread at Corsair's forum on the H100i: http://forum.corsair.com/forums/showthread.php?t=115981. Appears that Corsair link can tell you what the coolant temperature is for the H100i when the CPU is idle/load conditions. What is it on your system? According to OP of the thread, the H100i is sensitive to mounting. Might want to try the slight hand pressure on the cold plate to see if the CPU temps change while under load.]
 
#14 ·
Quote:
Originally Posted by GridIroN View Post

...

It's a Phantek EVOLV ATX, it's not a cold case to begin with, but my maintenance archive of apps is on a mechanical HDD. It was running the tests so it heated up most likely.
The Phantek EVOLV ATX case should give good airflow based on the picture here from Kitguru:

Did you mount the H100i at the top of the EVOLV or the front?
 
#15 ·
Quote:
Originally Posted by steveting99 View Post

As you've got test results for idle and load temps on the current BIOS, it would not be that difficult to flash the older BIOS and re-test.

If the test results are the same, check the coolant level on the H100i.

[Edit: See the following thread at Corsair's forum on the H100i: http://forum.corsair.com/forums/showthread.php?t=115981. Appears that Corsair link can tell you what the coolant temperature is for the H100i when the CPU is idle/load conditions. What is it on your system? According to OP of the thread, the H100i is sensitive to mounting. Might want to try the slight hand pressure on the cold plate to see if the CPU temps change while under load.]
Temperature of the water perpetually rose under Cinebench. CPU hit 92C, water hit 51C by the end.

I ran it again this time pushing on the block. No change.

 
#16 ·
Quote:
Originally Posted by GridIroN View Post

Temperature of the water perpetually rose under Cinebench. CPU hit 92C, water hit 51C by the end.
...
Coolant reaching 51 deg C under load and when ambient is 22 deg C is almost 30 deg delta. That's pretty large delta. Can you check the condition of the coolant in the H100i and it's level? If you've been operating the H100i for sometime, some evaporation of the coolant will occur over time. The the coolant level will drop.

If you have another CPU cooler lying around, try it and just check what the idle numbers are.
 
#18 ·
Quote:
Originally Posted by steveting99 View Post

Coolant reaching 51 deg C under load and when ambient is 22 deg C is almost 30 deg delta. That's pretty large delta. Can you check the condition of the coolant in the H100i and it's level? If you've been operating the H100i for sometime, some evaporation of the coolant will occur over time. The the coolant level will drop.

If you have another CPU cooler lying around, try it and just check what the idle numbers are.
There is no way to check the coolant level of an AIO cooler.

In any event, doing some more tests, it appears the H100i may be busted after all...the coolant continues to heat till about 52 C, and the CPU continues to heat with it. I just finished about 10min of 3DMARK physics test and whilst im typing this, the CPU is still around 70 C. It's cooling down very very slowly, which suggests the pump is broken, or the water isn't circulating for some reason.

Odd though, considering the pump reports 2237 RPM...
Quote:
Originally Posted by EMUracing View Post

The evolve top vents are very restrictive. if you don't isolate radiator, that hot exhaust air will recirculate back into the case and through the reservoir.

If you test with the side panel off is there much difference in temperature?
No.
 
#19 ·
Quote:
Originally Posted by EMUracing View Post

The evolve top vents are very restrictive. if you don't isolate radiator, that hot exhaust air will recirculate back into the case and through the reservoir.

If you test with the side panel off is there much difference in temperature?
Quote:
Originally Posted by GridIroN View Post

There is no way to check the coolant level of an AIO cooler.

In any event, doing some more tests, it appears the H100i may be busted after all...the coolant continues to heat till about 52 C, and the CPU continues to heat with it. I just finished about 10min of 3DMARK physics test and whilst im typing this, the CPU is still around 70 C. It's cooling down very very slowly, which suggests the pump is broken, or the water isn't circulating for some reason.

Odd though, considering the pump reports 2237 RPM...
No.
Scratch that. I had an testing app still running in the background. Water is down to 37. CPU is idling at 44.
 
#20 ·
Pull the H100 out of the case and test without any air restriction (leave block mounted on CPU). That will solve any possibilities that the case configuration is effecting performance. It is well documented in the watercooling section that a 240 radiator on the top of the Evolv needs to be isolated to work properly otherwise the vent restriction pushes the hot air down back into the radiator.

Other than that, could be an issue with the H100. I wouldnt suspect an issue with the IHS or needing a delid because the water gets hot. Since the water is getting hot, that means that block mount is ok, but either air is not flowing properly through the radiator, or there is low pump flow or low coolant (or both).
 
#21 ·
Quote:
Originally Posted by GridIroN View Post

...Water is down to 37. CPU is idling at 44.
So with the ambient at 22 deg C and the system at idle, the coolant temp is at 37 deg C while the CPU is at 44 deg C?

The delta (15 deg) between ambient and coolant looks suspicious, especially under idle scenario. Under idle, the CPU load should be in the low single digits and the CPU package power should be in the low teens. Under this scenario, the coolant should be a few degrees above ambient. Certainly not 15 deg C above ambient.

As suggested by EMUracing, temporary place the H100i radiator outside the Phantek EVOLV ATX case and check if there's any impact to the CPU.

If you have another spare CPU cooler lying around (air type is fine), swap out the H100i and check the numbers. Is it better, the same or worse than what you're getting with the H100i?
 
#22 ·
I spent last night and this morning playing around with the H100i in various configurations. I layed the case down and made sure the H100i's radiator was the highest point of the loop to check for air. I also shaked it, and reseated the whole thing a third time using different thermal paste, and my temperatures are again even worse. I'm idling at 60C now.

The H100i has one warm tube and one really cold tube. It almost seems to me as if one of the tubes is clogged or the liquid is messed up.

Either way, I've ordered an H100i GTX off Amazon and if it's it's not the cooler than I'll have to return it and look into replacing the 6600K if it turns out to be the CPU.
 
#24 ·
Quote:
Originally Posted by EMUracing View Post

But, have you removed it from the case? So it can breath free.

Shaking it will only create more bubbles, raising temps. If low on fluid, best orientation would be radiator vertical with tubes on bottom.
The EVOLV ATX has a top slide where the radiator installs. I slid the slide out, and confirmed the fans are blowing cold air over the radiator with no restriction. I left the slide out so the fans are blowing against nothing above them and there is no difference.
 
#25 ·
Quote:
Originally Posted by GridIroN View Post

The EVOLV ATX has a top slide where the radiator installs. I slid the slide out, and confirmed the fans are blowing cold air over the radiator with no restriction. I left the slide out so the fans are blowing against nothing above them and there is no difference.
Perfect way to test. With that done, next step is to test the new cooler when it arrives. It sounds like the water flow is sub par, but moving. Do you hear any gurgling as the pump runs?

Thanks for the description of the test.
 
#26 ·
Quote:
Originally Posted by EMUracing View Post

Perfect way to test. With that done, next step is to test the new cooler when it arrives. It sounds like the water flow is sub par, but moving. Do you hear any gurgling as the pump runs?

Thanks for the description of the test.
I hear nothing. Another guy at another forum said you should feel the pump vibrating. I don't feel anything really, but the PWM reports it's operating at 100% (2200 RPM)
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top