Overclock.net › Forums › Intel › Intel CPUs › 4790K drops clock frequency under heavy load
New Posts  All Forums:Forum Nav:

4790K drops clock frequency under heavy load - Page 3

post #21 of 29
Thread Starter 
Quote:
Originally Posted by Cyro999 View Post

If you apply a CPU load then the CPU will heat up to almost maximum temperature in a matter of seconds, certainly less than minutes and that is completely normal. Peak temps in the 70's and even 80's to low 90's will not cause thermal throttling.

This. Thankyou! BTW man do you think it could be a faulty CPU?
post #22 of 29
Thread Starter 
Quote:
Originally Posted by inedenimadam View Post

^ This. There may be some residual temperature creep that happens as the case heats up over prolonged usage, but hitting 70C within seconds is perfectly normal, and well in safe range.

I swear it sounds like VRMs overheating, but that board has a decent VRM layout, the chip is not heavily overclocked, and with the panel off and decent case flow... I fail to see how its possible.

Do you experience the same throttling when you leave everything stock? I get wanting to downclock a bit to keep the temps low, but wouldnt it be easier to just set a lower turbo multiplier and keep speed step?

Edit to add: your question about the north bridge: The north bridge connects the Cores to the RAM and PCIE, it used to be its own chip on the motherboard, but its functions have been integrated into the CPU die for several generations now. We still have a southbridge that handles USB, PCI, BIOS, and some other not so bandwidth heavy functions. My guess is that it is not actually dropping that low, and its just an error in reporting.

This certainly isnt a motherboard issue as I have switched board already but no use. When I am on stock my cpu never goes to 4.4 despite turbo on, even during prime95. but it goes to 100C instantly (which is a common issue with haswell according to google searches). And yea I also think its just an error of aida64. Because 30 Mhz is like extremely low. Is there any other program which allows north bridge monitoring? I am using HWinfo mainly but it doesnt monitor NB.
post #23 of 29
Quote:
Originally Posted by Symbfrk2 View Post


This certainly isnt a motherboard issue as I have switched board already but no use. When I am on stock my cpu never goes to 4.4 despite turbo on, even during prime95. but it goes to 100C instantly (which is a common issue with haswell according to google searches). And yea I also think its just an error of aida64. Because 30 Mhz is like extremely low. Is there any other program which allows north bridge monitoring? I am using HWinfo mainly but it doesnt monitor NB.

Aida64 does it best. It has its own monitoring suite built in, complete with remote server and customizable widget. Do some digging, its actually a pretty cool feature of Aida. OpenHardwareMonitor is another awesome program if you dont feel like using Aida. Just be aware, with more than one monitoring software open at once, it is not uncommon to have bad readings due to two programs polling the same data at the same time.

As has already been suggested, upgrading your cooling should be a top priority...but in the mean time, I would try reverting to stock and lowering the turbo multiplier to 40 (I think that is what you are aiming for?) instead of completely eliminating turbo.
Blue540
(14 items)
 
x99
(18 items)
 
B.M.O.
(10 items)
 
CPUMotherboardGraphicsGraphics
6800k ASUS X99-A EVGA 980 FTW ACX 2.0 EVGA 980 FTW ACX2.0 
RAMHard DriveHard DriveCooling
Corsair Vengeance CMK32GX4M4B3200C16W Crucial MX100 WD Blue 1TB EK Supremacy EVO 
CoolingCoolingCoolingOS
EK 480 PE EK 240 PE EK D5 Vario Windows 10 
MonitorKeyboardPowerCase
Samsung 49KU7500 (curved 4k) Corsair K65 NZXT Hale 90 V2 1000W Thermaltake X9 
MouseMouse Pad
Logitch G100S X-Trac Ripper 
CPUMotherboardGraphicsRAM
6400k msi A75MA-P33 Sapphire 6670 passive hynix 2 GB dual channel 
Hard DriveOSMonitorKeyboard
WD  windows 8.1 e-machine wireless 
PowerCase
king win 600w custom built 
  hide details  
Reply
Blue540
(14 items)
 
x99
(18 items)
 
B.M.O.
(10 items)
 
CPUMotherboardGraphicsGraphics
6800k ASUS X99-A EVGA 980 FTW ACX 2.0 EVGA 980 FTW ACX2.0 
RAMHard DriveHard DriveCooling
Corsair Vengeance CMK32GX4M4B3200C16W Crucial MX100 WD Blue 1TB EK Supremacy EVO 
CoolingCoolingCoolingOS
EK 480 PE EK 240 PE EK D5 Vario Windows 10 
MonitorKeyboardPowerCase
Samsung 49KU7500 (curved 4k) Corsair K65 NZXT Hale 90 V2 1000W Thermaltake X9 
MouseMouse Pad
Logitch G100S X-Trac Ripper 
CPUMotherboardGraphicsRAM
6400k msi A75MA-P33 Sapphire 6670 passive hynix 2 GB dual channel 
Hard DriveOSMonitorKeyboard
WD  windows 8.1 e-machine wireless 
PowerCase
king win 600w custom built 
  hide details  
Reply
post #24 of 29
Quote:
Originally Posted by Symbfrk2 View Post

This. Thankyou! BTW man do you think it could be a faulty CPU?

Probably not
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
post #25 of 29
Thread Starter 
Quote:
Originally Posted by inedenimadam View Post

Aida64 does it best. It has its own monitoring suite built in, complete with remote server and customizable widget. Do some digging, its actually a pretty cool feature of Aida. OpenHardwareMonitor is another awesome program if you dont feel like using Aida. Just be aware, with more than one monitoring software open at once, it is not uncommon to have bad readings due to two programs polling the same data at the same time.

As has already been suggested, upgrading your cooling should be a top priority...but in the mean time, I would try reverting to stock and lowering the turbo multiplier to 40 (I think that is what you are aiming for?) instead of completely eliminating turbo.

Thats a nice point there, I am running HWinfo and Aida at the same time, maybe that is causing the bad reading on north bridge clock, ill try again with just aida. I live in a 3rd world country and computers (high end ones) are really a luxury here so I usually import my stuff. I have ordered the cryorig H7 but it will take atleast 30 days to reach me. As far as turbo goes, id love to use it on 4.4 (Thats why I bought the K edition) but ill try the 4.0 setting too.
post #26 of 29
I'm having a very similar issue with my 4720HQ. Its almost definitely because of heating because it only happens when there is a large thermal load (CPU 50% or higher and GPU near 100%) and it happens after a period that becomes much smaller after the first occurrence. Also it happens when my CPU and GPU are not breaking 80°C and there is no thermal throttling indicator showing up in intel XTU.


Download intel XTU because it can let you now if you are getting thermal throttling and then you will know for sure if it is CPU thermal throttling or not. Also i noticed that it is possible for the CPU to heat up very quickly and even the 100ms refresh rate on hwinfo may not be fast enough to catch your CPU hitting its throttling limit.


And please let me know if you find a cause or resolution.
post #27 of 29
Thread Starter 
Quote:
Originally Posted by openglcg View Post

I'm having a very similar issue with my 4720HQ. Its almost definitely because of heating because it only happens when there is a large thermal load (CPU 50% or higher and GPU near 100%) and it happens after a period that becomes much smaller after the first occurrence. Also it happens when my CPU and GPU are not breaking 80°C and there is no thermal throttling indicator showing up in intel XTU.


Download intel XTU because it can let you now if you are getting thermal throttling and then you will know for sure if it is CPU thermal throttling or not. Also i noticed that it is possible for the CPU to heat up very quickly and even the 100ms refresh rate on hwinfo may not be fast enough to catch your CPU hitting its throttling limit.


And please let me know if you find a cause or resolution.

I disagree with 100ms part, because when the cpu throttles under load (non synthetic. gaming for example), it does so for a short duration and once temperature goes down a bit ( which only takes like a few seconds) it goes back to full speed and keep repeating this behavior. 100 ms might not be enough for catching every single throttling instance, but it certainly is good enough for catching it once.
And sure man, ill update if i find a solution.
post #28 of 29
Quote:
Originally Posted by Symbfrk2 View Post

I disagree with 100ms part, because when the cpu throttles under load (non synthetic. gaming for example), it does so for a short duration and once temperature goes down a bit ( which only takes like a few seconds) it goes back to full speed and keep repeating this behavior. 100 ms might not be enough for catching every single throttling instance, but it certainly is good enough for catching it once.
And sure man, ill update if i find a solution.

I mean that it might miss the temperature shift. We obviously know that your CPU is throttling. I'm just saying this because there were times when a slow update would show my CPU hitting 80°C maximum however in intel XTU I could clearly see that the processor was responding to its thermal throttling limit. I'm trying to help you identify your problem.
post #29 of 29
Thread Starter 
Quote:
Originally Posted by openglcg View Post

I mean that it might miss the temperature shift. We obviously know that your CPU is throttling. I'm just saying this because there were times when a slow update would show my CPU hitting 80°C maximum however in intel XTU I could clearly see that the processor was responding to its thermal throttling limit. I'm trying to help you identify your problem.

I understand your trying to help, i just stated what seemed appropriate to me. Anyway, I believe I have found the root of issue. It isn't thermal throttling (tried xtu just cause you said so), it's PSU... constant voltage fluctuations found on +12v (dropping down to 11.4 volt frequently) through multi meter. Since I dont have enough funds to buy a decent psu, ill have to wait for while before making the switch. will update once i do.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › 4790K drops clock frequency under heavy load