Processors
Intel® Processors, Tools, and Utilities
14536 Discussions

How best proceed with overheating i7-4790K?

REnso1
New Contributor I
281,699 Views

I have an i7-4790K in a GA-Z97MX Gaming 5 mobo latest F4 BIOS.

I used a Noctua NH-L12 cooler rated at 95W for cooling the processor rated at 88W in a Lian Li PC V354 with 4 fans, 2 in 2 out, case closed and case open, ambient is 27-30°C.

Temperatures in BIOS and memtest86+ were high so I decided to try stress testing and in Prime95 small FFT cores 1&2 overheated to 100°C using Core Temp.

I tried reseating the heatsink and renewing the NT-H1 TIM and opening the case but it made no difference. I have a photo of the contact pattern here.

When I tested using the OCCT benchmark I was unable to complete a test due to the processor overheating so I underclocked the processor to 3.6 GHz, disabled turbo and manually set vCore to 1.1v.

With an underclocked processor I was able to get a heating and cooling curve using the OCCT auto capture, to enable me to study the problem.

Even when underclocked the processor was reaching high temperatures, rapid fluctuations in temperature with work load suggest a bottleneck in the thermal pathway. When I tested with the intel retail cooler which came with the CPU the cooling was much less effective than the NH-L12 (even when underclocked taking just over a minute of OCCT to reach the 85°C cut off point see below) indicating the NH-L12 was doing a good job of removing heat, which meant the processor was making the heat or the source of the bottleneck.

I have discussed it http://forums.hexus.net/cpus/327593-4790k-overheating-nh-l12.html elsewhere. Advice was to contact Intel due to an absence of information relating to my retailer's testing procedures. I have asked about these but am still waiting for a reply.

So my question is how should I proceed from here? Does this qualify for an RMA? If so is it possible to negotiate this with Intel direct or do I have to go through my retailer?

I have done my best to make sure I am not doing anything wrong and I would be grateful for any pointers to any mistakes I may be making.

683 Replies
YNomi1
Beginner
6,399 Views

juanvaldino Inside the UEFI BIOS there is a Voltage Menu. There you can set VCore Max to the setting that cover your setup and thermal requirements.

0 Kudos
cgeor
Beginner
6,399 Views

but why not stay at normal voltage, because I have not tampered with anything everything is on auto , my friend has the same motherboard and he have normal voltage.....

0 Kudos
YNomi1
Beginner
6,399 Views

Juanvaldino have you checked that the bios version is the same (between your motherboard and your friends motherboard). I have 2 GA-Z87-UD5H and on both i had different results with f8 bios, different with f9 and the best results with f10c regarding the 4790K cpu.

0 Kudos
SBuck1
Novice
6,392 Views

I did an RMA through intel. My old chip was batch L421C113, and on all "AUTO" settings in bios would shoot to 100C at the slightest provocation. Small FFT test in Prime95 would cause the PC to shut down before I could react, temps shot up so fast. New CPU is batch L422B826. Passes the custom test posted last page at 58C. Prime95 Small FFT causes temp to near 100C if I don't cancel it, but better than instant shutdown. Idles at ~32C with CM 212 EVO cooler. Instead of locked the multiplier at 38 to be able to use my desktop I can leave it on AUTO and let the processor go to 4400MHz as per Intel specs. Overall I'm not hugely impressed with even this new chip, I feel it still has higher temps than it should, but that may decrease as the thermal paste breaks in, and it is now useable at stock settings.

0 Kudos
cgeor
Beginner
6,392 Views

Hello , i have intel i7 4790k and GIGABYTE GA-Z97X-UD3H , i make test stress on prime95 vers 28.5 during 10 minutes after i seen on cpu-z the core voltage is 1.430v , in bios all is default , how can I go back to the standard voltage? please

0 Kudos
cgeor
Beginner
6,392 Views

i update my bios and now voltage is 1.224 is ok ?

0 Kudos
YNomi1
Beginner
6,392 Views

I ve also just run v0.2. Passes fine with 47C being the higher core temp (core # 1). Room temp a bit lower this time 21C.

here is a screenshot:

Hope this helps.

Also my batch number is L430B875

0 Kudos
YNomi1
Beginner
6,399 Views

Also is there anyone from this thread that RMAd his CPU and got better thermal results. I am asking this because the thread is coming to a halt these last days...come on guys lets help each other a bit more. Do you have new results with changes to your setup, new cooler, rma'd cpu. Post them here.

0 Kudos
RKimb
Novice
6,399 Views

yianom wrote:

Also is there anyone from this thread that RMAd his CPU and got better thermal results. I am asking this because the thread is coming to a halt these last days...come on guys lets help each other a bit more. Do you have new results with changes to your setup, new cooler, rma'd cpu. Post them here.

Some of us are just waiting for a response from Intel. Maybe I'll live that long.

0 Kudos
SRene
Beginner
6,399 Views

I bought a new cooler, tomorrow it arrives and I'll test with it

0 Kudos
cgeor
Beginner
6,399 Views

i see now he have z-97x-gaming 3 and my is GIGABYTE GA-Z97X-UD3H , my last bios update is f7 but i trying with f6 to, and same voltage, it is big voltage or i can stay on this?

0 Kudos
cgeor
Beginner
6,399 Views
0 Kudos
cgeor
Beginner
6,399 Views

test on aida , after 40 seconds, i have high temp

0 Kudos
YNomi1
Beginner
6,399 Views

Your temps with intel thermal test are fairly normal (You should also provide me with ambient room temps). AIDA 64 & Prime 95 use complex instruction sets that give the CPU a hell of a day so 80C is also normal for that kind of stress. Anyway its only 6C above Tcase (which is something like max optimal operational temp).If your room temp is like 22-25 i believe you are OK. Please remind me the CPU cooling solution that you are using...

0 Kudos
cgeor
Beginner
6,399 Views

i have ZEROtherm ZEN FZ120 , i see on this site http://www.cpu-world.com/CPUs/Core_i7/Intel-Core%20i7-4790K.html http://www.cpu-world.com/CPUs/Core_i7/Intel-Core%20i7-4790K.html the CPU part numbers is different i have this one

  • BXF80646I74790K is a boxed processor without fan/heatsink

And my friend have this

  • BXC80646I74790K is a boxed processor with fan/heatsink (Chinese version)

It is difference voltage if we have different cpu part? when i buy it they tell me it is new revision of i7 4790k

0 Kudos
YNomi1
Beginner
6,447 Views

No i don't believe there is any difference between those 2 models.If you had the chance to use like a 100 4790K's you will see small differences between every CPU in terms of wattage and heat dissipation. The cooler you have is adequate (provided you use the fan and not use it passively), but you could see better results if you go hydro (like corsair h100i which i use ).

0 Kudos
idata
Employee
6,399 Views

Just received my replacement from Intel. Will let you know what the results are in a few days, I had no problems with RMA.

0 Kudos
KenF_Intel
Moderator
6,447 Views

Hello All,

 

We appreciate your continued feedback on this issue. We understand your frustration. We have been working to determine what is required for you to achieve the performance expectations you have for our i7-4790K processors.

 

Analysis of your feedback and our internal testing of certain motherboards have shown that some motherboard processor power settings are not set to Intel default specs and are thus requesting more power than expected.

 

The TDP for the i7-4790K is 88 watts. Many of you are reporting readings of 120 watts or more and a Core voltage of 1.3V or more. Remember higher voltage means higher wattage.

 

When using Intel's stock cooler, it is important to keep the wattage at or near the specified TDP (88W). By keeping within TDP you should be able to run at published Turbo Boost speeds (4.2GHz) for 4 active cores.

 

Screenshot 1 below shows settings that worked well on several motherboards that we used to replicate this issue.

 

We corrected the Core Voltage, Core Voltage Offset, Turbo Boost Power Max, Turbo Boost Short Power Max, and Processor Current Limits and completed a 30 minute XTU stress test with a maximum temperature of 74.1°C.

These settings worked on processors from different batches and on motherboards from different manufacturers.

Some important notes:

1. The settings available in XTU are dependent on what your systems BIOS exposes to the tool. The settings available may be different on your system.

2. The voltage, power and current settings that will work for you may vary.

Screenshot 1

The Intel® Extreme Tuning Utility (XTU) was used to monitor the wattage while performing the stress test.

You may need to turn ON the Wattage monitor by clicking the wrench icon just right of the graph section (see screenshot 2).

Screenshot 2

REnso1
New Contributor I
6,447 Views

First I want to say thanks ken for replying to the questions asked in the thread and providing a recommendation for settings. As the OP author I have marked it as the correct answer so people can find it quickly. Intel XTU is a nice utility and I appreciate the effort that has gone into making it available for Intel customers so thanks to Intel for that. I hope that by cooperating we can provide information that will be useful for all those trying to set up or test their I7 4790K's. So I would be grateful if we can work through the anomalies which crop up and work out what is going wrong and assist manufacturers in getting their BIOS working correctly.

I would like to get the ball rolling by saying that when I use the latest Intel XTU version with a Gigabyte GA Z97MX Gaming 5 BIOS F4 (latest) in Win7 Ultimate x64 and apply the settings above (Amps 105A, Watts 88W, Turbo BSP 110W, x44/43/42/42), the I7 4790K processor will run no faster than 3.6 GHz in the CPU stress test. That is with the voltage set in BIOS to "normal" with -0.05v offset because it is not available to adjust in Intel XTU. If I set the voltage manually in BIOS to 1.108v the processor will not run faster than 2.6GHz in the stress test.

With vCore settings reverted to the "normal"-0.05v value (because these produce a better if flawed result) and altering the manual settings in Intel XTU one at a time, I found the critical value in Intel XTU is the Processor Current Limiter, which has to be set at precisely 256 Amps or above in order to allow the processor to function above 3.6GHz in the stress test. With all other settings as specified in ken-Intel's screenshot and Amps set to 256 the processor will run the stress test at 4.2 GHz and reaches a maximum of 67°C in a 40 minute test under an NH-L12 with a maximum TDP of 81W.

This temperature and TDP does not reflect a current draw of 256A when other motherboards produce similar at 105A, yet setting the limiter even one click lower than 256 throttles the CPU to 3.6GHz when quad Turbo is set x41, yet the same 256 setting will also run x42 indicating this is not a result of a real power limit. It seems a curious coincidence that 256 is the critical value, when this is a recognisable binary number which has 8 zeros after the 1 and I am wondering if this suggests some kind of software error in the BIOS, or Intel XTU when working with this BIOS?

Can anyone explain this please? I would like to ask if this indicates a BIOS error and whether I should report it to Gigabyte?

KenF_Intel
Moderator
6,447 Views

BoolyBooly,

Thanks for sharing the settings that work with your system.

I definitely encourage you to report your findings to Gigabyte. Your settings may help other users who contact Gigabyte directly and it may help Gigabyte with future BIOS defaults.

0 Kudos
REnso1
New Contributor I
6,447 Views

Thanks for your reply ken, I have done as you suggested and reported the anomaly to Gigabyte who have, I want to add, provided the best customer support possible throughout this process (as have my retailer SCAN), so I have every confidence they will respond appropriately in due course.

0 Kudos
Reply