How best proceed with overheating i7-4790K?

[REnso1]

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.

[AOnIT]

Hi sampark,

(Edit: Quote below seemed to get chomped by the forum software):

"to say that ASUS, ASRock, EVGA, MSI, Supermicro all made the same exact mistake completely independent of each other is highly unlikely, wouldn't you agree?"

Certainly agree with that, this is definitely something more than board manufacturer issue. I was merely responding to Unconnected's insinuation that we're all fools for looking at and testing things other than concluding that everyone's problem is an IHS issue, which it clearly is not in every case. A lot of the process (including the tests you've run, which I followed with interest) were very useful to get to the understanding we're at now.

This would explain the drop of 20 degrees, which isn't common in a normal situation if you're simply swapping coolers.

Yes, in my (limited) experience 10-ish degrees is closer to the norm (at least, that's around what I got with an i5-3470 where I tried a budget aftermarket). But with a 4790K running proportionally hotter, while the IHS probably is a major contributor, it's pretty easy to conclude that the stock cooler contributes too, and just isn't up to the task in general -- unless perhaps you luck out with a "cool" 4790K? At this point I'd always recommend someone get an aftermarket even if they don't intend overclocking.

Cheers,

Ash.

[LCCF]

From KIT GURU review of Jan 15, 2015 re the Z97 motherboards,

http://www.kitguru.net/site-news/highlights/andrzej/kitguru-annual-awards-2014/3/. link to article

January 5th, 2015 at 12:34 pm - Author http://www.kitguru.net/author/andrzej/ Andrzej

This year's awards have been heavily influenced by Intel's various platform/chipset releases.

The Z97 was a bit of a damp squib initially. It promised so much, but as the http://www.kitguru.net/components/motherboard/luke-hill/asus-z97-a-motherboard-preview/ early previews started to come to light, it became apparent that not all of the planned technologies would be available on day one. Intel rushed to market, pulling in the launch date, and the result was a bit of a scramble from the various motherboard manufacturers to try and (a) differentiate their new offers from existing Z87 offers while (b) working to mask the missing parts.

Pity really, because the combined marketing machines of Asus, Asrock, Gigabyte and http://www.kitguru.net/channel/computex/faith/msi-displays-intel-based-world-firsts-at-computex-2014/ MSI were ready to go for it at Computex.

Let's get to the runners and riders in this category.

Z97 Motherboard Award

Arguably the toughest market of all, with very little to choose between the main rivals. With similar feature sets, capabilities and performance – price played as much of a part in this decision as anything else.

Our winner is a board that not only delivered well on the basics like M.2 SSD support, but which also managed to include Audio Boost 2, Killer NIC and XSplit GameCaster which lets you capture, record and stream your best gaming moments. And at a little over £100, it doesn't break the bank.

and as my post this morning (post # 644) indicated, ASUS just released another 3 frigging BIOS updates since i downloaded the 2501

[idata]

I have my new CPU. Pretty much the same than with the older. There is a slightly improvement. No throttling now, temperature reaches 90º after a couple of minutes, and last minutes temperatures is stable at 90~95º, with a peak of 98 for half a second. Indeed it´s slightly better than the older which was at 99~100º and throttling like there is not tomorrow for the last two-three minutes. This is using BIOS 2401 released in Jan.

I touched the heatsink and it was really hot, so in my case does not look like it is a inner thermal compound problem, but that the heatsink cannot manage to get rid of the heat from the CPU. I know about algorithms and logic, but I don´t know about air flow dynamics... and honestly, I am not willing to learn about it! So I leave it here.

larryccf I updated the BIOS with the latest releases. However as I suspected, there hasn´t been any improvement at all, the description for those three updates in Feb and March for Asus mobos indicates that is to support broadwell micros and better fan compatibility, the only BIOS update focused on system stability was released in Dec. And I am not hopping for a BIOS update solving this after 9 months of the release date.

I will think if I get a cooler or I leave this like it is, suffering occasionally of high temperatures. Really disappointed with Intel quality. For sure I will not get another just released, and will think if I get another Intel processor. What´s the point of getting something better than the competence (AMD) when if you want it to work normally you have to tweak it down?

See you out there with a cold beer!

Cheers,

K.

[BButt2]

OMG my head hurts. Yes I bought a new computer. Yes it is a I7 4790K, ASROCK H97 Pro4 Mobo. Yes it is at 99-100 degrees within a second of running any stress test. Yes, I have updated the bios and checked the seating. Yes, it has a stock CPU fan. I've loaded the case with fans, 2x140 up front, 2x120 top, 1x120 back (exhaust). Same, no change.

I've tried most all of the tricks and settings here to the point that it wouldn't even boot. It now boots again, but the same problem no matter what I do or any setting or CPU wattage throttling or Intel Extreme settings or anything else I've tried. I am fairly well convinced that the stock fan is not adequate, and it needs additional cooling somehow. The stock fan is going to be very hot no matter what you set it on.

I've read almost every post on this thread, and that seems to be the best answer.

So I bought a Coolmaster Hyper D92 fan. Arrives in 5-7 days from Amazon. I hope it works, about $44. shipped.

That's where I am - any suggestions would be appreciated but I think I've tried everything. I really don't want to read anymore anyway.

[OMysl]

Hey everyone,

I followed this thread and have the same problem with overheating i7 4790k. I have tried lowering and even undervolting the processor with little results. The processor still reaches 100 degrees in stress tests and has high idle temperatures despite very good cooling setup, several adjustments, checks of proper cooler installation etc - basically tried it all.

I've sent a warranty service request on intels website and today got a call from an Intel employee named Rafaela (or Rafael). She asked me about the details and all was going well until she told me, and I quote here: "it is completely normal that this cpu can run in temp up to 100 degrees and over, since its a very high performance processor meant to deal with such temperatures, as long as it doesn't shut down, it is completely ok for to run such temperatures" - I even asked her twice, repeating her statement and asking if I could quote her in the intel forum and she said yes.

When asked about the Intel stated operating temperature of 74 degrees, she said that its the maximum temp of the IHS outside cpu, so the actual core temps can be much higher.

Now am I crazy to think that it cant be possible for a processor to run such temps? Over a 100? Please correct me, but i feel like something is very wrong here.

[SPark48]

Wow... That's literally the first word that comes to mind.

I don't know about anyone else, but I have never heard of a desktop processor that anyone would say 100C is a safe temperature. Does this person realize that some plastics start to bend and melt at 120C? Perhaps short bursts at 100C would be OK, but prolonged exposure at this temperature would surely cause damage, right? I'm really confused that an Intel employee would give this as an answer. Perhaps she thought you were talking Fahrenheit? I'm just... wow.

Did she approve an RMA for you? Or did she just tell you to be happy with that?

[OMysl]

Hey sampark1980,

I was not convinced either, so I persisted in saying that I am not satisfied with those temps, the level of noice and that it did shutdown when reaching 100 degrees. Only then I was asked to run the Intel processor diagnostic tool, which I did before receiving the call and I told her that it passed all.

I asked if I could rma the provessor anyway and get another one since I am not satisfied with this one and she said I would need to send her a log from running the test.

Right now I am traveling and I wont be able to do that until next week. However, I am pretty tired after spending many hours over this issue, reading forums, tweaking bios and dissasembling and assembling my rig over and over. Dear Intel - I just want a proper processor.

PS: I do not want to get anyone in trouble, she was still very nice on the phone, however her statements did pose a worry to me.

Edit: well I hope she knew I was talking celcius, it did say in my request that I am from Europe and she called from a +44 number soo...

[GSing15]

Updated my BIOS today. Still, nothing. Played around with a different computer with different CPU, heatsink, motherboard, and memory, Same results. 100 C.

oneil366, i wouldnt bother RMAing. Save your money and time. I tried another 4790k that too with another setup.... it had the same issue.

[SPark48]

Actually gaganjit1, I did an RMA on my first processor that was overheating and the replacement CPU that I received is working fine, except with Prime95's Small FFT test. However, I do agree that it is a waste of money to have to pay to ship it to them. It's like rolling the dice. Some people get a good replacement, some people don't. My advice to you oneil366 is to request an RMA and at least see if the replacement CPU fairs better. What have you got to lose except about US$15 for shipping?

Ken, I know you are reading these posts, so can you PLEASE respond to the situation above with "Rafaela" who told oneill366 that 100C is safe for the 4790K? I could really use some clarification on that.

[GSing15]

Well, you can always try ...

[idata]

Hi,

My new processor after RMA has same issue, I would say has slightly improved but not a big thing. Just performing a test just after booting gets 90º but I do it after using the computer for a few hours, the 100º and throttling are there again. Unfortunately my RMA support guy, Meric, told me only one RMA due to this issue... Good thing, here in Singapore it is free of charge.

I am getting a better cooler today.

[KenF_Intel]

Hi Oneil366,

You're experience with our contact is not what I would expect. I'll correspond with you via Private Message to get some details so I can research the situation.

If you would like an RMA, please contact your local support center again.

[OMysl]

Hi again all,

After talking with Ken, I tried to change hte Vcore on my processor quite more than previously (before I only tried as low as 1.16V, no OC, 4.4 turbo).

VCore: 1.1V

My results: Around 80C on on Intel Burn Test standard stress (max 82 deg on high & very high stress lvl) comparing to 100C within a seconds on auto setting (1.276V max) and 92C on 1.16V.

I am too tired to redo the test with stock cooler, but I would expect the temps to be around 10C higher.

It runs stable on 1.1V, testing 1.08V right now - running 2C cooler on average.

Unfortunately idle temps are still high no matter the setting (manual 1.08V, 1.65V or auto): around 40C (fans running at 70%)

Setup:

OS: Win 8.1 pro x64Motherboard: Asus MAXIMUS VII RANGER, Z97

Bios: 2304

Ram: G.Skill DDR3 16GB (2x8GB) TridentX 2133MHz CL9 - no oc

Cooler: SilentiumPC Fortis 2 XE1226 (SPC097)

PSU: Be Quiet! POWER ZONE 750W 80PLUS Bronze (BN211)

CASE: SilentiumPC Aquarius X90 Pure Black (SPC098) 3x 120mm intake fans, 2x 140mm out

All fans running 100% during stress.

For testing I used intel burn test 2.54.

I am tired right now and I think i will stick with this one with no RMA, at least for now. Maybe a delid in the summer or once I get brave enough

Hope that helps...

PS: Thanks to Ken for reacting really professionally to a bad support received via phone.

[JMesz]

Didnt read the whole thread but I did make it to page 22, but I'm a "me too" with the overheating i7-4790K cpu.

My board is also a Z97X-UD5H. I have tried many things to deal with this overheating issue. Here's a list of what I have done.

* Read and applied the information in the document for boards that run hot. However, my CPU will only reach 2.77GHz with the supplied settings.

* Cleared CMOS (a bunch)

* Reflashed current BIOS (F8) & cleared DMI, & loaded defaults

* Re-seated the cooler. Still using stock cooler for now.

* Installed the Tuning Utility & Adjusted my settings to match the information given in earlier pages.

However, I managed to work my temps down to the upper 70s, and for the heck of it tried the benchmark option in the utility. My score is 566, FAR below everyone else with the same board. I can marginally increase it to 577, but my temps are mid to upper 90s. I cant undervolt anymore else I cant boot and use the machine. No OC is going on, the headroom for me isn't even close.

I did get the system to maintain a mid 70s overnight with BOINC running on all cores without issue. I may go back to those settings, but if my performance is going to suffer that bad, I have to figure out why. $330 CPU after all.

My CPU info is:

Code: BX80646174790K

Batch: X439B539

Is my CPU a bum CPU? I got it from Amazon, so I don't know well they'll accommodate frequent CPU swaps if there are still a bunch of "bad" ones out there. Anymore thoughts or updates on this problem?

[PSchw5]

I am troubleshooting a hot 4790K as well. This is a computer is not OC'd and is really only used for general internet activities. It hits 100C and throttles instantly on any stress test and often runs into thermal throttling under standard PC tasks such as YouTube videos. This is my fourth Intel build in recent times, but the only one with issues. The only thing unique about the build is I gave the stock CPU cooler a shot. I do not feel that is the issue because it is at stock clocks and the heat-sink does not get even warm to the touch when the thermals are showing 100C and how quickly the CPU jumps to 100C (basically instantly confirmed by the graphs).

Screen shots of specs etc follow. Latest Bios, reset multiple times, max voltage of about 1.1V (hard to tell with how quickly it throttles), XMP disabled, multipliers at 44/44/43/42, etc. Considering an RMA, but not clear that it will resolve the issue. Looking for input (Ken?).

[JMesz]

My temps are still really high, but I did a complete reinstall with Windows 7 Premium yesterday and my benches are much better. I'm at least in the right area of performance using the XTU benchmark, even my SETI crunch times are a little better.

Obey, you can click the little wrench icons and show the throttling and core voltages. I'm able to get my machine to run stable by setting the Vcore in BIOS to 1.095v, but it wont run 4 cores at 4.4GHz (BSOD) but runs fine at 4.2, and it helps to keep temps "low". See how low you can take your core voltage, it helped me quite a bit. I get a few issues when I set my core at 1.090v or lower (mostly no post) 1.093v boots fine and runs fine under full load.

I had to tweak the TDP else my cpu wouldn't run more than 4.0GHz, essentially turning off the Turbo Boost.

Here is my settings. Note that my voltage and power max & current max are set in BIOS, and are unchanged in this app.

Friendly Disclaimer: I'm not responsible if you hose your rig, you know your system better than I do so don't blame if my setting croak your CPU.

[PSchw5]

Appreciate the reply and tips. It looks like it was set to peak at 1.2V, so I'm walking it down slowly now.

Also, nice disclaimer, haha.

Edit: Ended up at -0.031V offset (1.091V) stable through the Intel Stress Test, but it only gave me another 1 or 2 seconds of non-throttled 4.4GHz. Looks like I have a dud.

[SB18]

Hello everyone,

I read the whole subject, and I am glad I found people with exactly the same problems as me following the purchase of a "4790K" I feel less alone.Like you, I mounted my system and like you I launched test programs to "test" my new CPU. I also to put my PC "POWER OFF" in 1-2 seconds using "Prime95 v28.5"My system:- Gigabyte Z97X-UD5H- CPU 4790 K- Watter cooling (Aquaduct 720 XT)After hours and hours of research here and on the net, I finally found "my" solution to have a CPU that works as it should have run home.Here's what I changed:- I removed "IHS" CPU http://forums.anandtech.com/showthread.php?t=2261855&page=4 Delidded my i7-3770K, loaded temperatures drop by 20°C at 4.7GHz - Page 4 - AnandTech Forums- Thermal paste used as "Liquid Coolaboratory pro" directly on the "DIE"- Though cobbled together for adjusted my "Watter block" on the modified CPU.- Forced a Vcore in the BIOS to 1.191 V (not "Auto")- Put in a XTU "offset", that way my CPU does not exceed "1.2710 V" in full load.Now I can run without problems (finally) a "Prime95 v28.5" with 4 cores at 4.405 GHz, without "throttling" :-) and finally very reasonable temperatures.Sorry for my bad english, I remain available to answer questions and to later screenshots if you think it's useful.

Thank you.

Swiss_Intel

[BCicc]

swiss_intel, that's a solution that voids warranty. I believe people should not be forced to take that road in order to have a well-functioning product.

[SB18]

Yes, I agree ... but I wonder how could I managed to prove that the Intel CPU that I purchased works ... but evil.

When I read the problems to "organize" an RMA (or need to prove it), I just found a solution that works for me.

... And since I computer (over 25 years) ... I have never returned a CPU under warranty, they are obsolete long before

EDIT

- with "Prime95 v28.5" no problem with very good temperature.

- with "LinX 0.6.4" no problem with very good temperature.

- with "LinX 0.6.5 - 11.3.0.004" reboot my computer 1-1.5 seconds http://www.xtremesystems.org/forums/showthread.php?201670-LinX-A-simple-Linpack-interface/page36&s=dcdfb620c96cc7896e6c340de62165fa LinX - A simple Linpack interface - Page 36

[NClar]

I'm hoping someone could offer some advice.

I recently installed a 4970K and immediately ran into temperature issues. I'm fairly experienced with building PC's and I'm sure the thermal solution was not the problem (I reapplied with new TIM, twice, just to be 100% sure). Despite having correctly implemented the bios changes outlined by intel as a result of this thread- my temps are passing 85° under real world load, with an open case and an ambient between 15° - 20° (it's winter currently).

I came to the conclusion that the CPU must be faulty and initiated an RMA with the vendor. I opted for a refund rather than replacement, because I intended to replace the faulty 4790K with a 4790S- hoping its lower TDP (65 W) would be the safer option.

The problem is, I just read http://www.techspot.com/news/59696-intel-s-series-4790s-power-performance.html this article which concluded that the S series actually uses just as much power, under real world conditions- thus generates just as much heat. But maybe they weren't testing these under the right conditions? ie. Maybe they weren't aware of, so didn't address the incorrect bios defaults issue covered in this thread?

So I need some advice about what's the smartest way to move forward. should I…

1) Replace the faulty 4970K with another 4970K, and hope it doesn't suffer the same issue? or

2) Replace it with the 4790S, assuming the conclusion in the above article is ill informed?

Anyone's thoughts on this would be very helpful.