Thermal sensor issue i7-7700k?

[CK7]

I have a brand new build; EVERYTHING NEW. i7-7700k is running at stock speeds. I have the RAM set to XMP for DDR4-2666. Motherboard is Asus Maximus IX Hero Z270.

I have found that the i7-7700k reports a momentary (a second or less) temperature spike +25 > 35 degrees Celsius anytime a program is opened, a webpage is opened, a background app runs etc. The temperature blip cascades through the cores in random order; not the same every time. This causes my heatsink fan to constantly cycle up and down. Temperatures otherwise report as steady, normal increases. Peak temperature under Prime95 blend test is 71 degrees Celsius.

Attempted solutions:

I have re-installed my heatsink and thermal paste with no change.

I have tried to manually set my fan speed in the bios. The only setting that avoids this issue is setting the temperature / fan at a constant (and loud) 80-100%. I've tried PWM and DC mode.

I have found a few user reports elsewhere on the web, all reasoning that it's just the way it is. I don't accept that. Opening a folder or browser should not spike temps +30 degrees. Not only is the fan cycling annoying, it puts undue stress on my fan; possibly shortening its lifespan.

What's the answer, if any? RMA?

[AElib]

The only thing i could suggest is for those who have extremely high temperature/spikes to return the CPU before it's too late.

[idata]

I hope we all have a support case with Intel on this issue ... and that we keep prodding them directly, as well as on this and other forums.

[DBux]

A little summary of my history with this issue:

1- I bought and made myself a new PC (I'm doing this regularly from, at least, 1992); all new components, air/fan cooler and 7700K, Mobo Gigabyte Z270 Gaming7, 16Gb 3200Hz RAM. NO OVERCLOCK.

2- Quickly I noticed strange spikes because the fan went crazy, from top to idle in little seconds, only opening a new tab in explorer or notepad. I have never seen something like this before (and I mounted enough K processors before).

3- I made hard stress test and reached 100C with ease (Mobo alarm +90 was always beeping). That's when I began to be afraid.

4- I discovered this thread and explained my case. Also I read all valuable info from other complaints.

5- I changed my air/fan cooler by water cooling system and tried changing bios settings. Things went better, no more 100 (now max 90C with hard gaming).

6- I'm still having temp spikes only opening notepad, from 35C to 70C more or less. Water goes the same mad as before, but less noise at least.

7- Intel has no response then, while I'm waiting, I'm getting used to this spikes (and mad cooling system, etc.) because now it doesn't reach more than 95C in most hard stress test, and, well, this temp are under CPU official limit.

8- So... yes, I can life with this issue. But I don't want, I don't want because I don't want to give up in front of Intel's contempt, dismissing all people here. It's OBVIOUS that something isn't right with this Cpu then we want, at least, an answer from Intel. And, of course, if this spikes are 'a new feature' of K processors then we want to know it (because, in my case, I will never buy a K processor again).

Then TGrable (and other doing same thing), please... we know that this issue can go better with better cooling system/pressure/paste, with deliding, with bios... but we are not here to listen advices like punishment for doing bad things; avoid to blame us like newbies, because perhaps some are, but most not. We must use our forces to complain and to force Intel to give us an aswer, not to sow the doubt in front of that overwhelming issue.

[idata]

"Then TGrable (and other doing same thing), please... we know that this issue can go better with better cooling system/pressure/paste, with deliding, with bios... but we are not here to listen advices like punishment for doing bad things; avoid to blame us like newbies, because perhaps some are, but most not. We must use our forces to complain and to force Intel to give us an aswer, not to sow the doubt in front of that overwhelming issue."

I absolutely agree that we need to keep pressure on Intel; however I also think that it is very useful for us to tune our systems to get the best results we can, and to share these results with each other, as TGrable is suggesting. This will demonstrate to Intel that we have tried everything possible, made sure that we have good cooling solutions, that we have checked that the coolers are correctly mounted, that we are not running the CPUs at voltages that are too high, etc ... in other words that we are serious and not a bunch of guys who are whingeing because we have nothing better to do.

At the very least we should all run our systems at stock frequencies (memory included) and default BIOS settings with Prime95 small FFT, with AVX, and report the results here (with our system details); and we should all show a graph of the temperature spikes when starting applications from idle. If we all did that (and reported the results directly to Intel), then we would be helping each other decide whether or not our CPUs are really seriously out-of-spec compared to the average, and we would be giving useful information to Intel.

If we also did this with our best overclock effort then we would be giving even more useful information.

I'm away at the moment, but when I get home this weekend I will do this (although I have already given most of this information in this thread).

Cheers

Robert

[DBux]

Of course, I'm adressing to TGrable but is nothing personal, is only about the 'sow the doubt' attitude. All Tgrable (and other people) advices were very valuable and useful (thanks a lot), but I think isn't necessary to sow the doubt everytime and on everybody who is complaining here. Is clear that could be a lot of reasons that make issue vary or be worst, but I think that what is clear is that THERE IS AN ISSUE. Then don't doubts, don't try anytime to sow doubt if is an issue or if is user's fault.

[TGrab1]

DanBux wrote:

Of course, I'm adressing to TGrable but is nothing personal, is only about the 'sow the doubt' attitude. All Tgrable (and other people) advices were very valuable and useful (thanks a lot), but I think isn't necessary to sow the doubt everytime and on everybody who is complaining here. Is clear that could be a lot of reasons that make issue vary or be worst, but I think that what is clear is that THERE IS AN ISSUE. Then don't doubts, don't try anytime to sow doubt if is an issue or if is user's fault.

It isn't about "sow the doubts."

It is two fold. I want people to use the resources available to them to try everything they can on their end. This is the two fold part... because they both help themselves in the event it improves and also they help the case to intel by showing the effort that was used to try and get satisfactory temps.

I have never once said that there isn't an issue. I agree there is one, I have constantly agreed on this and also given my opinion to the cause.

Now reading your second to last post Dan, I also have a question for you. I am looking at the steps you took, but did you at anytime look at your voltages and maybe think of tweaking them. Your temps are either the results of a horrible chip (TIM wise) or voltages that are too high for the stock clocks.

There are still the motherboards out there that are either not updated to current on the bios or that are still having issues with auto voltage... so if you are running at stock clocks and are seeing voltages over 1.2 or so I would think you could reduce them some more to really help on temps... Like my previous picture I think at stock volts I was running 1.14-1.15 volts under max load. So even if you can reduce your volts to 1.18 at stock that should net you a good improvement in temps... I fear with the temps you are reporting that you might be in the 1.25+ range.

So check if you can and let me know. It might be your chip having horrible thermal properties but it could be voltage and that is a mostly easy fix.

[DBux]

Haha, you are doing again, TGrable, you are incorrigible! But I don't want to fall in your trap! (I'm joking). (Yes, you're right, I am moving about 1.2v; but some stress tests gave me bad results when less, and I decided to use the first stable setting I found avoiding to reach too much over 90C. In future I will make more tests and will try more settings - also I will overclock at some point -, but this is not the main complaint and purpose of this thread! Thank you anyway).

[DFara]

Hello people,

it is definitely the same problem presenting itself here.

My system:

Motherboard: ASUS PRIME Z270-A

CPU: Intel core i7-7700k (downclocked to a manual vcore of 1.185 volts, currently running at 4.6Ghz Turbo)

CPU cooler: Noctua NH-D15 - the secondary 140mm fan is swapped with a 120mm Noctua NF-F12 fan due to RAM heights. The spare fan is fixed on the top of the case above the heatsink, pulling.

RAM: 2x8GB of Corsair Vengeance DDR4 3200Mhz

Main drive: Samsung 960 EVO NVMe SSD 256GB

GPU: Gigabyte GTX1070 WF2OC

PSU: Corsair RM750X

Case: Corsair Carbide Series 200R

A 10 minute stress test in CPU-Z peaks at an approximate 82-84 degrees Celsius (the temperature reaches the upper seventies very fast, in 3-5 seconds), depending on ambient thermal values and system uptime. Don't even get me started on heavier stress tests. All five of my fans inside the case are running 100% at this point, of course. Idle temperature is about 45 to 50 degrees, that although with having all the vdroop and speedstep features turned off. All these results are accomplished by reducing multiple voltages, including vcore, dram, vccio and the system agent (vccsa); before that, it was even worse (upper 80s to 90 degrees under full load). I have already eliminated the possibility of mounting the heatsink the wrong way; I reapplied the thermal paste twice, then the third time, I went ahead and completely dis- and reassembled the kit, and reseated the CPU itself as well. I tried two different thermal compounds. None of these experiments resulted in relevant change regarding heat output.

I can confirm that this CPU is indeed very sensitive to the slightest of change in workloads, temps are all over the place just opening a browser window.

One more thing to note is that the heatsink itself stays relatively cool as the thermal readings go up. Everything points to an incredibly weak rate of heat transmission inside the CPU, or best case scenario, as the thread title suggests, we should be throwing the numbers out of the window.

Don't get me wrong everybody, my system is perfectly stable and I guess I'm alright with it, but it is still quite a frustrating experience overall no doubt. I was obviously hoping to ensure a hefty OC but instead I can't help myself to be constantly on the lookout for something to go wrong, running more or less on just the stock settings. I fail to believe that 95 degrees can be safe for this unit, at least until somebody explains the reasoning behind that notion... not one single previous processor model by Intel suggests that that's a healthy boundary. What is the actual difference here compared to previous high-end chips? Please give us a response, Intel.

[idata]

Well, add another to the list. Well, three actually since I am on my third 7700k. Yep, three. I originally upgraded from a 6600k on a z170 board and lets just say temps were not good. At stock voltage/clocks the thing was pushing 90s and I had to under volt it just to try to keep temps in check. The second chip was better, but not by much and was still displaying temps at stock settings that seemed way too high. So, I though maybe it's my cooler. Therefore, I went and bought a $150 280mm AIO hoping it would solve the issue, but nope. Temps were better overall, but the random spikes were pretty much just as high. Fortunately, the store I bought the original chip from has excellent customer service and allowed me to exchange it for yet another 7700k. And I even swapped out my z170 board for a solid z270 board thinking perhaps it was a bios issue with the z170 mobo.

Long story short, while this third chip on the z270 board is displaying better temps than the other two, most of the time, under heavy load it still pushes 90 degrees, and under light loads, while just browsing the web, for example, will still spike into the 70s. And that's the odd thing, even under a constant load, all three chips have displayed the tendency to go from perfectly fine temps(60s and 70s even under heavy load) to all of a sudden randomly spiking out of nowhere into the 80s or 90s, just to drop back down. And whats even odder, these spikes will happen at all different voltages and clocks. For example I've had higher temp spikes at times while undervolted and underclocked than I have gotten at stock voltage/clocks.

And that's under a consistent load. Whereas, as mentioned, going from idle, or near it, and being asked to perform a light task spikes the chips into the 70s. I also notice the clocks tend to be all over the place, which might be part of the problem, at least for the spikes when asked to go from low load to performing a light task. The chips almost never stay at low clocks but constantly jump around and then go full speed at the lightest task. right now sitting on this page my chip is jumping around from 800mhz to 4.5ghz. And just opening a new tab causes it to jump to 4.5ghz with temps spiking into the 70s.

I really don't know what to do at this point. A forth chip doesn't seem like the answer since a forth is likely to display the same behavior. And while delidding is the solution you'll hear from most people, that is not something I am willing to do, nor should it be necessary. Besides I even talked to a guy who did delid his, and is on a custom loop, and yet just opening his browser still causes his temps to spike by 40 degrees. lol

Anyway, as I said, the store I purchase my hardware from has excellent customer service and they were even willing to let me go with another processor last time. I could have bought a AM4 mobo and a Ryzen chip if I wanted to, but I said no I wanted to stick with the 7700k, since on paper at least, it is the best CPU for what I want to do. But at this point I'm starting to wonder if I made a poor decision.

[FT2]

I had the same problem.

This is my Setup:

Motherboard: MSI Gaming Pro Z270

CPU: Intel core i7-7700k boxed

CPU cooler: Noctua NH-D15S

RAM: 2x8GB - Corsair DDR4 2400Mhz

SSD: Samsung 850 EVO 500GB

GPU: actually none (integrated i7 ....the gpu is currently in delivery)

PSU: Be-Quiet Dark Power Pro 11 650W

Case: Fractal R5

So at the first time I started the computer the biosshowed me a idle temp of 30°, so far so good, then I updated the bios to the newest version installed win10 with some programs like firefox and coretemp. After installing the basics I wanted to try Prime95 26.6. The temps were insane high (about 80+) and the voltage was 1,3 -1,35v. I thought it's my fault and i did something wrong with the thermal paste, so i reapplied the tp and tried another paste. Result: temp +5 in idle. I removed that tp again and reapplied the same type of tp like my first try. result: temp -5, same results like first time building that pc. So i decided to try the voltage settings in bios. All were set to Auto. I'm not a big fan of overclocking so i did the voltage of the cores to 1,160 (i think it could be less). and set all other values to the default values. After a reboot I had no more temp spikes. The idletemp is low in the 30s. I ran prime95 about 2 hours and the temp was lower than 60 degree. The Corespeed is at 4500ghz (in windows energyoptions, maximum performance). The watt usage while running prime95 was constantly at 70W. Surfing in the web and watching youtube 4k (on a 4k TV) video didn't spike the temp anymore.

So i think there's a problem with the "auto" voltage/oc settings in the chipset/mobo/bios. I hope the guys from intel and the mobo manufacturers will give a bios updates to fix this "auto fault settings".

Maybe this helps somebody!

Best,

Florian

[ACozo]

I lowered cpu temp to 1.2V. It ran ok for 2 days and then it started to hang at certain applications. Only solution was to set the temp back at auto. Application is running fine again.

My temperatures reach 97°C. I asked for a replacement chip. DHL picked up my i7 7700K last thursday. It is one week now and I still don't have the replacement chip. It's taking a long time.

[MaryT_Intel]

Hi Florian 1992,

We are definitely looking into everything. The CPU is operating within parameters but no one likes the heat spikes so we are investigating all options. We don't have any new news to share yet but when we do, this will be the place. Thanks,

Mary T.

Support Community Manager

[idata]

Hello Mary,

The spikes are certainly an issue for a lot of people (mainly because of fan noise) ... however for many of us the real issue is heat and not spikes. The fact that replacing the TIM between the die and IHS with a better heat transfer compound fixes this issue points inescapably to Intel having used a poor quality TIM. There really is only one solution to this problem, and that is to replace the TIM.

My particular issue is that I do not want to delid the CPU, thereby voiding the warranty. I also do not particularly want to have to spend $50 or more getting the kit required to delid the CPU safely (but that is a side-issue).

One acceptable compromise, for me, would be for Intel to extend the overclocking warranty to cover a CPU delid for the i7 7700K. This warranty can of course exclude physical damage to the CPU package.

I think that this is a reasonable suggestion. Nobody is going to delid the CPU for the fun of it, and as you can tell from this forum, this is a genuine issue for us and one that is not going to go away. It may be that Intel is doing some fine research into a solution to improve the heat issue going forward, but this does not solve the problem for customers who have already purchased this CPU.

It is also going on 5 months since this issue has been reported, and still no solution from Intel ....

Robert

[ASušt]

I personaly think, that intel is not researching anything. The final words were said a long ago.

These spiky CPUs will probably live with their thermal issue for some time.

So intel can sit сalmly, not worring about mass returns.

Intel has it's x299 platform comming, so all interest is gathered there.

[NStau]

Same here, brandnew machine, CPU cooler is ramping up and down like crazy:

Asus Z270G Gaming, newest BIOS 0906

Intel Core i7 7700(non K) with stock cooler

RAM: 4x8 GB 2133 Kingston

SSD: Samsung 960Pro 512GB

GPU: 1080Ti Founders Edition

OS: WIn10 x64

I'm going to replace the cooler with a Be Quiet Dark Rock 3 tomorrow and simply connect the fan directly to the PSU.

I'm not getting any other side effects (haven't measured the temps yet), but the fan cycles are driving me mental.

To be honest, this is really disappointing...

[JBárt]

Please can anyone from Support Team of Intel write any information? For example: We are still investigating it. We are ignoring it. Or magical word: very soon. Don't worry, I want only know any informations (good/bad/still nothing). It's very frustrating.

[YRenk]

I am waiting same answers too and i guess everybody is waiting ( But INTEL still keep silence

[AElib]

I am not waiting anymore and i don't even give a shit what Intel will say ( if they say something? )

[YRenk]

Even became 5 monthes and nobody tell us nothing . Maybe u r right , they doesn't care ... I think , this will be my last INTEL cpu .

[idata]

i bought i7 7700k 2 days ago and i got spike too. idle 38 degree sometimes hit 68 degree only one sec or two sec and case fans going crazy. it's annoying. anyone try RMA for that issue ?

[TGrab1]

JT13 wrote:

i bought i7 7700k 2 days ago and i got spike too. idle 38 degree sometimes hit 68 degree only one sec or two sec and case fans going crazy. it's annoying. anyone try RMA for that issue ?

Intel has been more than happy to issue and RMA for these chips, but a new chip will not solve the "spike" issue. It will just mean you are playing the silicon lottery on the quality of your next chips thermal efficiency.