Question Attempting to reverse overclock

QCube

Honorable
Jan 25, 2014
106
0
10,690
ASUS Maximus VII Hero
Intel i7-4790K

I've been having a lot of troubles recently playing in games whereby I freeze and go straight to blue screen. I've checked over my system and noticed that I'm overclocked which right now, I don't need that kind of power to my components. The i7 is peaking at 4.4GHz in some cases where I am just idle.

AI Suite shows that I'm 10% overclocked which I hadn't realised. I've gone into the uefi and I've loaded system defaults for everything to which it's still overclocked. The only options I have on the Uefi is Auto, Manual and XMP.

Noticing that, AI Suite also shows my DRAM Freq is 1330Mhz but my ram is 2333Mhz.

Any help or solutions?

Thanks!
 
4.4Ghz is Intel's max (single-core) boost frequency. If you don't want to use Intel's turbo boost you should be able to turn it off in BIOS.
If your CPU can't even boost without becoming unstable then you've got some thermal issues that should be resolved. What's your cooling solution? Case airflow? Has the case (heatsinks, fans, etc.) been cleaned recently?

Maybe try turning off/disabling the AI Suite and resetting your CMOS? Then just load up HWiNFO64 to check system specs...?
 

QCube

Honorable
Jan 25, 2014
106
0
10,690
4.4Ghz is Intel's max (single-core) boost frequency. If you don't want to use Intel's turbo boost you should be able to turn it off in BIOS.
If your CPU can't even boost without becoming unstable then you've got some thermal issues that should be resolved. What's your cooling solution? Case airflow? Has the case (heatsinks, fans, etc.) been cleaned recently?

Maybe try turning off/disabling the AI Suite and resetting your CMOS? Then just load up HWiNFO64 to check system specs...?

Was hoping someone might of stumbled across the actual setting in the uefi because I tried to hit it into my XMP profile but it then got stuck into a boot loop but cleared that with the CMOS reset.

I'm running with the H110i GTX, my GTX 980Ti SC died on me this week so It's all been cleaned out. Small amount of new thermal paste (MX-4) , checked over the VRMs, I've had to buy a GTX 1060 Super just to get me by for now and that's brand new.

I haven't come into a blue screen other than when I'm under load for longer periods (Only when gaming). I've been monitoring temps whilst I'm in game, CPU doesn't go anything above 65° and GPU is a nice cool 42°. My next best option is that I disable the overclocking, given that I don't need the power but the stability for the time being (Computer is also used for work purposes)

Cheers!
 
XMP just overclocks your RAM. It doesn't do anything with your CPU.

The setting will be called Intel Turbo boost, or some combination of those three words, and will be a disabled/enabled setting. I don't have that MB so I don't know specifically what it's called or where it's at.

Try this - Don't change anything yet. Load up HWiNFO64 (sensors only) and leave it running while you game or stress the system with something like RealBench. You don't want it to get to blue screen but want to see if any of the limits get hit under the 'Performance Limit Reasons' section in HWiNFO64, so you'll need to be checking it periodically.
 

zx128k

Reputable
It could be that his bios is reset to default and because the XMP profile is not loaded. His ram is at 1333Mhz. This is just a complete guess. I believe resetting to defaults again would not have changed that.

Default memory speeds are, https://ark.intel.com/content/www/u...-4790k-processor-8m-cache-up-to-4-40-ghz.html

Memory Types DDR3-1333/1600, DDR3L-1333/1600 @ 1.5V

Max Turbo Frequency 4.40 GHz

I would reset AI Suite to default or uninstall it.
 
Last edited:

QCube

Honorable
Jan 25, 2014
106
0
10,690
XMP just overclocks your RAM. It doesn't do anything with your CPU.

The setting will be called Intel Turbo boost, or some combination of those three words, and will be a disabled/enabled setting. I don't have that MB so I don't know specifically what it's called or where it's at.

Try this - Don't change anything yet. Load up HWiNFO64 (sensors only) and leave it running while you game or stress the system with something like RealBench. You don't want it to get to blue screen but want to see if any of the limits get hit under the 'Performance Limit Reasons' section in HWiNFO64, so you'll need to be checking it periodically.

Thanks for letting me know about this software, hadn't used it before! I've actually got NZXT Cam software because i have LED strips in my case that I prefer to customise and with that, It shows clock speeds, fan speeds and temps. This shows the CPU jumping from 800Mhz to 4.4Ghz, fan speeds going up and down in tandem with clock speeds. Just want to get the overclocking removed for the moment as It also sounds like I have a fluctuating jet engine sat next to me :\.

Performance limit reasons:

0c2d91f082e2f9a5804d354de2efeb6c.png
 
Last edited:

QCube

Honorable
Jan 25, 2014
106
0
10,690
So I just tried to load a different game (Rainbow six siege) and see if that had any different outcome on performance limit reasons and I wasn't surprised that I had the blue screen again before I could go and check again.

Stop Code: WHEA_UNCORRECTABLE_ERROR

Not sure where I'm to go next really. I've just bought a GTX 1060 Super and I can't play any games just yet.
I'm really missing my GTX 980Ti :(
 

QCube

Honorable
Jan 25, 2014
106
0
10,690
Your CPU had a 'thermal event' and went down to 800Mhz which is its LFM (low frequency mode).
This has nothing to do with your graphics card. You have a thermal issue with your CPU.

I removed the CPU recently and was in good condition as was all the pins. Thermal paste I've changed to Mx-4 with a small 'pea' amount. I use a Corsair H110i GTX
Haven't come across thermal issue before with an Intel processor. Could this be a M/board issue or even a fault in sensor readings?
 
I removed the CPU recently and was in good condition as was all the pins. Thermal paste I've changed to Mx-4 with a small 'pea' amount. I use a Corsair H110i GTX
Haven't come across thermal issue before with an Intel processor. Could this be a M/board issue or even a fault in sensor readings?
Doubtful.
Are you checking your temps with multiple programs? Some may be inaccurate.
HWiNFO64 can monitor.
 

QCube

Honorable
Jan 25, 2014
106
0
10,690
Doubtful.
Are you checking your temps with multiple programs? Some may be inaccurate.
HWiNFO64 can monitor.

I ran HWiNFO64 whilst gaming and temps never exceeded 65° C or so. Nothing crazy too high that I've come across.

I've also got iCUE installed (For mouse, keyboard, Mousemat & H110i GTX) and I've got it set to pump LEDs to red if CPU temps was > 90°C - I know this isn't the most reliable method but was an option so that I don't have to tab out of a game to check temps/clock speed when I could end up blue screening at any moment. It hasn't exceeded and gone to red at any stage.
 

QCube

Honorable
Jan 25, 2014
106
0
10,690
Your CPU had a 'thermal event' and went down to 800Mhz which is its LFM (low frequency mode).
This has nothing to do with your graphics card. You have a thermal issue with your CPU.

XMP just overclocks your RAM. It doesn't do anything with your CPU.

The setting will be called Intel Turbo boost, or some combination of those three words, and will be a disabled/enabled setting. I don't have that MB so I don't know specifically what it's called or where it's at.

Try this - Don't change anything yet. Load up HWiNFO64 (sensors only) and leave it running while you game or stress the system with something like RealBench. You don't want it to get to blue screen but want to see if any of the limits get hit under the 'Performance Limit Reasons' section in HWiNFO64, so you'll need to be checking it periodically.

WHEA UNCORRECTABLE ERROR is a common BSOD error that may occur due to a faulty hardware component. However, the most common cause is low voltage to the processor and overheating CPU due to overclocking.


So I need some help on controlling my CPU as It's becoming rather unusual and I believe it's why it's leading me to get blue screened.

This was on idle earlier and I decided to let it shutdown to cool off instead of seeing what would further happen.
uKtn2o3.png

This is what currently I'm having over my CPU - Is it normal for the frequency to keep flucuating this much? dropping to 800-900Mhz straight up to 4Ghz+? This is also at idle and see the graph in the picture which shows the CPU frequency. Notice the CPU fan is at 100% also which sounds ridiculous given that it's at idle. No overloading programs loaded or processes that's taking up 100%.

RjooCEE.png

Any help would go a long way in solving my issue!
Appreicate all the help regardless.
 
A couple things don't make sense. You said that, during gaming, your CPU never gets above 65ºC but then you are showing a pic of your CPU at 80ºC at idle...?

You have a CPU thermal issue. The fact that your CPU is clocking down to its LFM proves this. How much voltage are you feeding the CPU? Have you verified that the AiO is working?
 

QCube

Honorable
Jan 25, 2014
106
0
10,690
A couple things don't make sense. You said that, during gaming, your CPU never gets above 65ºC but then you are showing a pic of your CPU at 80ºC at idle...?

You have a CPU thermal issue. The fact that your CPU is clocking down to its LFM proves this. How much voltage are you feeding the CPU? Have you verified that the AiO is working?
CPU temps have only recently began going to 80 degrees at idle. Earlier this week I was gaming with no cpu going this hot (at least not what I monitored whilst gaming)

If you are running at stock clocks and your turbo modes are disabled in the BIOS:

Check if your pump on your AIO cooler is working. Your fans may be spinning, but pump failure can still occur, thus not circulating the cooling fluid.

What would be the best way to determine pump failure? I have the h110i GTX. Is there software to test this? Still have yet to find the turbo disabled function in the bios menu. Asus UEFI is straight forward but can't seem to have any option of not overclocking to 4.4Ghz.

Thanks!
 

zx128k

Reputable
CPU temps have only recently began going to 80 degrees at idle. Earlier this week I was gaming with no cpu going this hot (at least not what I monitored whilst gaming)



What would be the best way to determine pump failure? I have the h110i GTX. Is there software to test this? Still have yet to find the turbo disabled function in the bios menu. Asus UEFI is straight forward but can't seem to have any option of not overclocking to 4.4Ghz.

Thanks!

My pump has a tachometer that can be plugged into one of the motherboard fan headers.
 

Karadjgne

Titan
Ambassador
You have iCue and a Corsair aio. You can get rid of Cam, it's doing the same thing as iCue can/does and multiple reporting programs can be a cause of confusion on the pc and lead to errors.

If you have reset bios to optimal factory default settings, any OC or potential OC is gone. The only thing that'll override that is the Turbo-V in Aisuite which is software so gets applied after bios and windows loads. The only thing I've ever used is FanXpert, which was awesome for fan control in earlier generations.

I'd dump everything except the FanXpert from Asus. (FanXpert can be used standalone), dump cam, get iCue working for aio control, then reset bios and cmos.
 

QCube

Honorable
Jan 25, 2014
106
0
10,690
I've factory reset my bios on every single option, everything is running on auto but still overclocks and sends my CPU 85 degrees.

url]
jeyaWFa.jpg

url]
m43s3hN.jpg

url]
1ep6I67.jpg

As you can see the Asus extreme tuning utility doesn't give me an option to disable, only auto, manual and xmp. Resetting all options to default in bios doesn't change this.

I don't actually use AiSuite 3, but seen it on a post elsewhere, whereby there was an ability to default apply settings to stop an overclock but that was just another option I tried.


I've been having nightmare with my computer this week. It started over a week ago, my computer just died midgaming and I immediately smelt burning. After I switched it off, it would not turn back on again. After Id taken everything apart, I thought the M/board had died, I checked all the vrms and everywhere and found nothing. I took the GPU out and it fired right up.

I've just purchased a Gigabyte 1060 Super and plugged that in and it's got me back up and running. I took apart my GTX 980 Ti and I can't find anything wrong with it. No burnt capacitors, chip in good condition, I cleaned it all down and bought a new heatsink & fan to go on it, applied new thermal paste. I tried putting that back in today and it wouldn't turn on with the gpu installed so I think it's dead.

Since my computer crashed and I smelt burning, I've been having all these problems with my CPU which I never even had before and also blue screening (probably down to it throttling)

You can see in my pictures attached above, 85 degrees in the bios menu which I watched it climb to that temp. You also see that everything is set to auto. Cpu voltage is 1.088V which is also, Auto.

You have iCue and a Corsair aio. You can get rid of Cam, it's doing the same thing as iCue can/does and multiple reporting programs can be a cause of confusion on the pc and lead to errors.

If you have reset bios to optimal factory default settings, any OC or potential OC is gone. The only thing that'll override that is the Turbo-V in Aisuite which is software so gets applied after bios and windows loads. The only thing I've ever used is FanXpert, which was awesome for fan control in earlier generations.

I'd dump everything except the FanXpert from Asus. (FanXpert can be used standalone), dump cam, get iCue working for aio control, then reset bios and cmos.

I have cam installed because I have the Hue+ installed in my case. iCue is running because I have K95 RGB, M65 Pro RGB, MM800 RGB mouse at and of course the h100i GTX.

I have done a bios reset, cleared CMOS with a battery pullout and still the same.

I'm wondering if my EVGA GTX 980Ti SC ACX 2.0 is in working order but my CPU problems are what prevents it from turning on with the 980Ti installed. I'm obviously gutted to be losing my GTX980Ti SC and replacing it with a 1060 Super. Huge difference and feel like I've just lost £500. The GPU board looks brilliant, I've even gone over it with a magnifying glass and found nothing.

I've lost all hope right now, felt like I had a real nice rig but I've gotten so disheartened with all these problems over the past week.
 
If your CPU climbs to 85ºC in BIOS and you've 'smelt burning' you've definitely got an issue.
Have you tried buying a cheap CPU cooler to see if it's your AiO? Also, get rid of ALL overclocking utilities/softwares, for now - complete uninstall.

You may be facing a failed CPU or motherboard. After testing with a new cooler, your next step is to purchase/borrow a new motherboard or CPU to test.
 

zx128k

Reputable
85c in the bios menu at idle for the cpu is a cooling failure. If I had a full custom water loop and get 85c at idle with stock voltages. If I then smelt burning I would check the pump.

The other story could be that the AIO could have leaked, took out the gpu. GPU is dead and the cpu is now overheating. If your system would not boot with the original gpu after you smelt burning that's a bad sign. One that you should not ignore.

This is one that needs troubleshooting by someone that knows what they are doing and has parts on hand. You could be lucky and replacing the AIO will fix your cpu overheating.

I had a Koolance water cooling kit a long time ago, the pump pcb went on fire. So I replaced the cooling system without checking the PSU was still working correctly. The PSU after a few weeks died with clouds of white smoke. So I then replaced the PSU. A nice short story at least.
 
D

Deleted member 14196

Guest
the cpu could be heat damaged at this point if you smelled burning. toasted cpus don't perform well, if at all
 

Karadjgne

Titan
Ambassador
To reset to factory settings is F5, don't do it manually as there's multiple settings that are in bios that are not shown by the UI.

Smelt burning is bad. Not just from gpu standpoint, you'd be best off getting your nose involved as that particular scent doesn't go away in a hurry. It could have originated anywhere, from psu to gpu to mobo to back of mobo to pcie. Which all can have an adverse reaction on the cpu.
 

QCube

Honorable
Jan 25, 2014
106
0
10,690
85c in the bios menu at idle for the cpu is a cooling failure. If I had a full custom water loop and get 85c at idle with stock voltages. If I then smelt burning I would check the pump.

The other story could be that the AIO could have leaked, took out the gpu. GPU is dead and the cpu is now overheating. If your system would not boot with the original gpu after you smelt burning that's a bad sign. One that you should not ignore.

This is one that needs troubleshooting by someone that knows what they are doing and has parts on hand. You could be lucky and replacing the AIO will fix your cpu overheating.

I had a Koolance water cooling kit a long time ago, the pump pcb went on fire. So I replaced the cooling system without checking the PSU was still working correctly. The PSU after a few weeks died with clouds of white smoke. So I then replaced the PSU. A nice short story at least.

Sorry, just to clarify the burning smell was over a week ago and I haven't smelt it since. I was gaming when the computer shutdown itself. I tried to turn it back on and when I stood up, I immediately smelt burning so I hit the power switch and shut everything down. The room smelt of burning electronics that even my girlfriend noticed straight away.
Since then, I haven't come across the burning smell, but all these problems have occurred now.

I bought a new heatsink and fan assembly to replace my old one on my 980Ti which I plugged in and installed today. Computer would start with the 980Ti installed. I removed the 6pin and 8pin power connectors from the GPU and the computer actually started with the obvious "please power down and plug in power to the graphics card". Tried it again and wouldn't boot after putting power to the GPU. As I mentioned before, it was in really good condition with new thermal paste, no damage to capacitors or soldered parts.

The unfortunate thing is I don't have access to testbenches or friends that live locally for me to be able to test or borrow pparts to help me eliminate further.


To reset to factory settings is F5, don't do it manually as there's multiple settings that are in bios that are not shown by the UI.

Smelt burning is bad. Not just from gpu standpoint, you'd be best off getting your nose involved as that particular scent doesn't go away in a hurry. It could have originated anywhere, from psu to gpu to mobo to back of mobo to pcie. Which all can have an adverse reaction on the cpu.

I've factory reset all bios settings using the F5 option, not sure if I mentioned above but this also didn't do anything.

I realised burning electronic smells don't just vanish so I've put my nose to work for the last week and found nothing. I've checked over the M/board with magnifying glass, under the VRMs, under the ROG logos. Taken out the CPU, cleaned all that down, still no burning smell on either the front or back of the M/board.

Checked my 2 HDDs and my SSD over but these haven't corrupted or have any burning smell to them.

I have the EVGA G2 Supernova 850W F/Modular psu. I've had it 4 years so it's out of warranty so I went to the extent of opening this up, checked out all of the capacitors, nothing blistered and no burning smells.

Checked over the RAM but these seem to be in working order.

Ideally need to test my 980Ti on another bench to perhaps see if there's damage to my mboard (although my 1060 Super works fine) but perhaps the psu can't supply enough power to my 980Ti and won't switch the computer on.

I never suspected the AIO to be a culprit but that's a possiblity if the pump has failed. I loaded up the computer earlier and tried to see if it would stabilise in games but I checked temps and CPU is hitting 92 degrees which is where I got to the point of shutting it all down and walking away in pure frustration lol.

Edit: I don't believe there's been a rad leak at all. What's the best way of checking a closed loop???
 

Karadjgne

Titan
Ambassador
The Evga G2 850w has a 7 year warranty. Get ahold of Evga if there's any doubts about whether the psu is good or not.

There's no way to test a closed loop cooler for a leak, the only way to diagnose a leak is by visible evidence, which is usually staining on the motherboard, corrosion around the edges of the pump base, spray drop patterns if the pc is quite dusty etc.

A pump consists of 4 basic parts, the motor, the diaphragm, the microfins and the pcb. It only takes 1 to fail and the pump is bunk, and most assume it's the motor.

If the pump doesn't work at all, it's usually something on the pcb is toast or motor is frozen. If there's a vibration, the motor is working, but the microfins can be clogged or the diaphragm is toast, either results in little to no flow. The last is lack of coolant. After years of use, the coolant starts breaking down into its composite elements, H²O turning into hydroxides and Oxygen. Oxygen molecules are tiny, and will eventually start bleeding through the rubber hoses. If enough of that happens, the pure liquid level goes down, upto the point where there's insufficient liquid to get pumped throughout the loop and cpu temps go way up. This usually happens @ the 5-6 year point.

With that cooler, it's large enough in capacity that absorbtion rates will not be a problem. So both hoses should be the same temp, ± 2-3°C at best. Since one is hot and the other colder and it's self evident, I'd be inclined to believe there's a flow problem. Either lack of coolant or some sort of blockage to the microfins and thats preventing adequate flow. The X62 should be relatively new, I'd be looking to see if it's covered under Warranty still and give a shout to NZXT.