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.

[SPark48]

gaganjit,

Considering your gradual increase in temperatures rather than your temperatures shooting up instantly to 100C+ instantly, I think your issue is that your CPU cooler isn't sufficient enough to keep your temperatures down. This is further evidenced by your TDP fluctuating somewhere between 80-90W. I know you said that getting another cooler was not a solution for you - is that because you're weary of using MORE money to buy an aftermarket cooler? Whether it's fair or not (that you are forced to pay more money to buy an aftermarket cooler because the stock cooler that comes with the processor from Intel is not good enough), an aftermarket cooler would most-likely solve your issue. There was a post that larryccf put up a couple of pages back that has a link to some great coolers you can buy. I would give those a try first. Good luck!

Sam

[idata]

Hi...

sampark1980

How does affect, or how should affect, the environment to get those extremes temperatures (assuming a correct cooler installation), or for the stock cooler being able to absorb the heat generated? Meaning by environment what case are you using (dimensions, etc...), your room temperature, available space within the case, graphic card used and whether it is being used at the moment, etc...

May all these factors increase the overall CPU temperature up to 20º (from 75~80 to 95~100) and a better cooler would help to decrease the temperature? I think that considering a hot environment, within reasonable limits, should not increase the overall temperature so much. I cannot give precise figures, but maybe the difference could be 5 ~ 10 º?

gaganjit1 what is your rig set up?

[GSing15]

keridito, I run a MicroATX box.

PSU : XFX TS 550WMemory: 10 GB Samsung DDR3 1333GPU : Depends 660Ti or Integrated GraphicsMoBo: ASRock Z97M AnniversaryHDD: 2.5 TB Total WD ( 1+1.5TB)Ambient room temperature is 21 C , and motherboard temperature reports to be between 25C to 32C max. CPU 95C max. GPU 75C max.

[idata]

My case is similar to yours, in term of increasing of CPU temp. It is not immediate to reach 95 - 100º but it takes a few minutes. In my case I am using a MicroATX (Urban SD1), although my room temperature is higher (~29º very humid Singapore), and also mobo´s ( ~35º). Also using stock cooler. Larryccf was also using a small case, and Noctua Cooler helped to decrease CPU temperature. I totally agree with you that we should not have to buy a aftermarket cooler when we are not aiming CPU overclocking...

[GSing15]

keridito, consider us equal then. Mine also takes about 4 minutes maybe to go to around 95C, but it hits it alright with the stock cooler. I just wish intel gave me a better cooler, or a warning that i would have to buy one. Their response is poor. Literally, this chip is like an overclocked i7 4770k without any effort on cooling. Not even, optimized for cooling. I'm now stuck with an overheating CPU. I need a $60 cooler to keep this at normal temps, Intel's effort: create a $15 cooler that will allow it to go over 100C. That's just BS. It's like buying 100mbps internet and being told that only for 5 minutes i can get that speed, after that, you get 20mbps. Plain BS.

[idata]

But then, if our case is not exactly the same than those CPUs getting 100 in a couple of seconds, and considering that there are people getting normal temperatures with the stock cooler, is it just the environment as I mentioned? If it is, just a well ventilated small case makes 20º difference, being necessary to get a better cooler? If it is not, then may be this a different problem with this processor?

My CPU is currently under RMA, Intel Malaysia just received and it is under evaluation. I will update any news...

[GSing15]

I highly doubt it's an airflow problem. I've tested 2 4790k CPU's, both got 95C. 1 of them intel tested getting under 80C or something. I mean I've ran stress test with open chassis and to be honest, the inside of my case was barely warm. Seemed like a few degrees over room temps, that's it. I think it could very well be a cooler issue....or i suck putting thermal paste really really really bad (from the likes of it, i think im okay with application).

[LCCF]

guys - i'm going to repeat what i said above, about my board's BIOS being the main culprit - think about what i posted - a 20C drop in temps while rendering video files, after installing the latest BIOS. And don't forget i had initially seen a 5-6C drop from simply un-installing the ASUS AI Suite III software or utilities. BTW, others on the tomshardware forum indicate all the mobo mfgrs software utilities create conflicts in the BIOS. Yesterday i ran a 3 hour rendering of a video file, and it actually shocked me to see the highest temp (in RealTempGT) had been 61C, with temps running 51-55C approx 95-98% of the time. For the hey of it I went back to the ASUS support page for my Z97M-Plus board, and noticed that in the first 21-22 days after release of the board, ASUS had released 3 BIOS updates, one of them 3 days after releasing the previous BIOS "update". If that doesn't indicate sloppy rushed work in BIOS scripting, i'm not sure what does.

And i'm not picking on ASUS here, i suspect all the mobo mfgrs are rushing product out onto the market. And considering the computer market is so price point focused, i suspect we'll never see a "mercedes benz" mobo mfgr where they charge for quality, including proper evaluation during pre-production testing, but deliver a properly finalized / quality mobo.

Interestingly, on other forums, i've twice seen reports that at least one mobo mfgr (Gigabyte) has in the past, shipped mobos with quality components (capacitors etc) on the initial product release, then changing to lower grade components after all the reviewers have tested / reviewed their product. And Gigabyte isn't alone in this "business model".

and this latest BIOS that i just installed was only released 10.5 months AFTER the Z97M-PLUS was first released. I'm starting to think the mobo mfgrs rush product out to market, without fully or proper evaluation testing, not wanting to lose any sales to competitors' boards, thinking they'll correct any issues in the BIOS later.

I haven't noticed any drop in voltages, my core voltage was showing 1.121 before, and still hitting that value, and while my benchmarks have not fallen that much - profiles i had saved in XTU as "954 benchmark" are now posting a benchmark of 949 - 950. I really think it's just conflicts in the BIOS that have been remedied. I'd noticed stability wasn't that great as every 10-14 days XMP would be dis-abled, power values were jumping back up to 4,095 & 1.023 watts, fan speeds were changing with no rhyme or reason, "wake on lan" settings were changing and a few other values changing. Think back to that post where someone mentioned their BIOS had a multiplier of 8 (or was it 11) when you punched in new power values for "turbo boost" max and "turbo boost max short burst" into the XTU settings. I don't know enough about software script or BIOS writing to say definitively, but values changing unpredictably / spontaneously would seem to be from conflicts rather than just incorrect or high values.

the part that irks me the most, is i just ordered a larger case ($109) that would accept a taller cooler ($67) and the BIOS remedied the situation before i had a chance to move my mobo into the new case. And prior I spent close to $50 buying additional fans for the case. All expenses i could have avoided, as well as the headache of reading a, at the time, a 31 page thread on this forum, looking for a solution.

The lesson i've chiseled into my memory, is next time i buy a mobo, i will avoid buying one that hasn't been on the market 10-11 months, at a minimum, to let the mobo mfgr get around to doing what should have been done before releasing the mobo.

The blame thru out this thread has been focused on intel, for the most part - early on the focus seemed to be "de-lidding the cpu" to correct the overheat problem, then it wandered but no one "fix" brought about the change in temps that the latest BIOS did for my board. A 20C drop while running a fairly intense video rendering file for extended periods is not an insignificant temperature drop. BTW, even now my BIOS is not conflict free - i now get a cmd or dos screen that flashes on the screen for a micro second every 15-20 minutes, not all the time but a good percentage of the day. And it only started doing this with the latest BIOS, so there's something still amiss in the BIOS.

One way to confirm my focus on mobo BIOS would be for one of the micro-atx board users above buying an ASUS Z97M-Plus board and running it with the 2501 BIOS, but that's a sizeable expense (they're currently running appro $100 on amazon), and seeing if they see the same drop i have. But as of right now, the only item i fault intel is for supplying that stock cooler with the obnoxiously noisy 80mm fan on the devil's canyon cpu - even not supplying a fan would have been a better judgment call. But even that is not an "all condemning" mistake.

I've only posted this rather lengthy response as everyone's focus seems to be on the cpu - think about something, it would have been extremely foolish and short-sighted on ken-intel's part to have sent gaganjit1 a cpu that he hadn't actually tested and known to be performing properly, and yet when gaganjit1 saw the same temps, the blame is still focused on intel. Intel may be a "greedy corporate" entity, but they are not stupid, nor hire staff that would be that foolish.

IF your cpu isn't trying to fry itself running the programs you normally run (i'm assuming folks don't normally run prime95 24/7) then live with higher than normal temps until your mobo mfgr gets up to speed with a decent BIOS update. 

[SPark48]

keridito,

To be sure, ambient temperature and things like good case cooling and airflow are all vital parts of maintaining a good CPU temperature. I'm not arguing against that. However, I'm sure most of the people on here have enough tech-knowledge to realize this and their overheating problem persists in all environments. Personally, I build and sell computers as a hobby and I've built my fair share of computers. I'm also a bit of a perfectionist so my cable management has to be perfect with everything color-sleeved, case fans have to be perfect, I even paint all my parts with engine enamel so everything matches with motherboard. I mean, if you look at my builds, I'm sure you'll agree I'm building very nice computers. I mostly use Corsair or Bitfenix cases, although I customize every single one with my own emblems and colors. For me, I'm fairly certain that it's not an issue with case size or airflow or anything outside of the CPU itself being too hot. I am also fairly certain 99% of the people on this thread have as much, if not more technical knowledge than I do. My main issue is that many of the people I'm building computers for are requesting the 4790K processor, but I have to discourage them and convince them to get another processor, OR, even worse, I build it for them and they are now complaining of overheating with the 4790K and thinking I did something wrong with the build. It's very frustrating when you're trying to build top-of-the line computers yet you can't use the latest i7 processors because of this issue. (And yes, the 5820K and 5930K also have overheating issues.) I don't mean to offend you at all when I say this, but I don't think talking about what kind of case you are using is helping the problem. It's not like we're saying "Oh my, case size! That's it! Why didn't I think of that???" It's painfully obvious by now that it is something with the CPU itself, and not case size, airflow, or cable management. We are on here, angry with pitchforks, because we have already exhausted every alternative option we could think of ourselves, yet the problem is not solved.

larrycf,

Again, I have no doubt that ASUS has a hand in your issue. What I'm saying is that if it was an ASUS issue, only people with ASUS motherboards would have this problem. You're focusing on YOUR system and so you may think ASUS is the issue, but there are people on here with motherboards from every manufacturer that are having the same problem. Doesn't that strike you as too much of a coincidence if Intel has nothing to do with this overheating issue? So you're saying that somehow every manufacturer managed to put these outrageously high settings into their BIOS completely independent from each other? I find that highly unlikely. And, as I mentioned above (and you can confirm with threads just like this one) that the 5820K and 5930K also have these same issues with overheating. I find it much more likely that Intel is the issue and that the processors Intel provided to manufacturing companies to test and build their motherboards were different from the ones that made it to production.

[LCCF]

sampark - you missed my point. As the computer industry is so price point fixated or focused, and sales volumes can easily run in the millions, mfgrs are rushing product out to the market BEFORE doing proper evaluation pre-production testing, ie not finalizing the BIOS as thoroughly as they should, figuring it's smarter (and it is) in their business model to get the marketing campaign started, with all the racy "titanium extreme carbon fiber" terminology worked into their marketing campaign, and capture all the market share they can, figuring they'll correct any BIOS issues later. Problem is, they're all forced to follow their competition's lead, whether it was ASUS that started this trend, or gigabyte etc, doesn't matter.

a 20c drop in max temps was a hellacious drop, after all the effort of changing coolers, improving air flow, re-applying thermal paste etc. Even had i known enough to have paid attention to the indications of stability issues with my BIOS, i wouldn't have thought i'd see that kind of difference.

go to your own mobo mfgr's support page for your board and see what they've got in the way of BIOS update history. What i saw on ASUS's (and posted a shot of) said a lot to me - their product was rushed out onto the market.

ken-intel has been mentioning BIOS settings since early in this thread, and i'm wondering / realizing how about the conflicts in "rushed out the door" BIOS (updates included). Again, think about the replacement cpu he sent to gaganjit1 - do you think he'd be foolish or short-sighted enough to have sent a cpu that wasn't performing properly? And yet when gaganjit1 tried it, he saw the same temps - what is the common factor but the environment the cpu was placed in, ie the mobo and it's bios

fwiw

[SPark48]

gaganjit1,

I agree with you 100% that you absolutely should not be forced to buy an aftermarket cooler to run your processor at STOCK; you should be able to use a STOCK fan! However, you don't need to spend $60 on a new cooler. There are a ton of $15-$30 range coolers out there that have more than enough cooling power to keep your temperatures normal at stock. Also, I'm not sure if you're nearby a Frys, but right now they have a Zalman CNPS20LQ liquid cooler on clearance for $25. That's not a typo, it's twenty-five dollars for a liquid cooler. It is a clearance item so probably an open box item, but still... It's well worth that price. If you can't go to Frys, I've had very decent cooling experiences with some of the cheaper Cooler Master CPU fans on Newegg and Amazon. They look almost like the stock fans, but larger heatsink and fan. Those should be more than enough to keep you in good temperature ranges if you're not going to overclock. Those are like $10-$15. If you don't mind paying a little bit more, the Cooler Master Hyper T4 costs only $28 on Amazon right now.

I know none of this is the answer you're looking for. You shouldn't have to spend MORE money to get your CPU to work correctly. However, I just think it would be better than having a processor that you can't use properly.

Good luck and let us in on your progress.

[SPark48]

Actually Larry, I believe it is you who missed the point. You mentioned gaganjit1, you do realize he has an ASRock motherboard right? Not an ASUS? So... his ASRock motherboard has a problem because of ASUS? You're blaming ASUS for the OP's overheating even though he has a Gigabyte motherboard? OK... haha

[SPark48]

Let me ask you something Larry. Let's say there are 3 cars. A Ford, a Chevy, and a Toyota. You own the Ford. Let's say just for the sake of argument that Company X provided the SAME exact engine to all 3 of these cars. Let's say that engine blows out on all 3 cars if you hit a speed of 80mph. Your engine blew out at 80mph. You are blaming Ford because you have a Ford. Yet the Chevy owner had the SAME PROBLEM! The Toyota owner?? SAME PROBLEM! Who is at fault? Ford? Chevy? Toyota? Or Company X?

[MM12]

All of this is anecdotal.

I keep getting email updates on this thread, and shake my head at the goose chase people are still led on.

PLEASE, do yourself a favor - when your CPU is running at 80-100C, FEEL your heat sink/fan. It aught to be very warm if it's functioning. If your heat sink/fan is not considerably warm or hot, your problem has nothing to do with BIOS' or settings or voltages.

Getting a 20c drop by switching coolers is unheard of. The stock heat sink fan is sufficient. Between cooler changing, something was mounted differently and now within the CPU, there is better contact between the IHS (The big metal piece on top of the CPU) to the CPU.

I RMA'd my 1st 4790k and had an identical problem with the second unit until I delid it, used better thermal compound, and properly installed my CPU cooler.

2 4790k's, 1 motherboard with the same bios and everything, same problem with both chips. Intel even blamed my motherboard. (Asus Gryphon Z97) The problem was not the contact between Heat Sink and CPU, it was from the CPU's lid to the CPU within.My cooler was cold to the touch while my CPU was cooking away at 100c.My CPU now runs at extended max load around 70C

[LCCF]

sampark - not sure if you're speed reading my post or just not reading it in it's entirety -

".....As the computer industry is so price point fixated or focused, and sales volumes can easily run in the millions, mfgrs are rushing product out to the market BEFORE doing proper evaluation pre-production testing, ie not finalizing the BIOS as thoroughly as they should, figuring it's smarter (and it is) in their business model to get the marketing campaign started,.............and capture all the market share they can, figuring they'll correct any BIOS issues later. Problem is, they're all forced to follow their competition's lead, whether it was ASUS that started this trend, or gigabyte etc, doesn't matter.

one easy way to identify or isolate it as the cpu would be for one of the micro-atx users above to purchase the same mobo i'm running, with a nearly finalized bios and see if the problem still exists. If the overheating issue disappears, then it's confirmed as not being a defective cpu

in case, newegg has some open box z97m-plus boards for $101 http://www.newegg.com/Product/Product.aspx?Item=N82E16813132130R&cm_re=z97m-plus-_-13-132-130R-_-Product Open Box: ASUS Z97M-PLUS LGA 1150 Intel Z97 HDMI SATA 6Gb/s USB 3.0 Micro ATX Intel Motherboard - Newegg.com

[LCCF]

if all three mfgrs are rushing their engine managment software to race product to market, then yes, it would be all three mfgrs, ( ford, chevy and toyota)

again, we've now got one motherboard, the z97m-plus that is running a cpu that was overheating and now is not -

[LCCF]

unconnected: you said "...2 4790k's, 1 motherboard with the same bios and everything, same problem with both chips...." That could also be used to confirm my point, re mobo bios

i don't doubt your delidding brought temps down, but pls save me from rereading a 41 page thread and remind me how many degrees it brought those temps down.

when i see my rendering temps drop 20C from a bios, i'm a little shocked - if conflicts in the bios are not there to cause the heat, then there wouldn't have been need for delidding.

btw, there's a recent bios update # 2205 that asus just posted, dated 2/17, on your board's support page, and if you haven't tried it, the asus ez updater was probably the easiest update approach i've used - works from within windows - took a total of 5 minutes including the reboot

fwiw

[MM12]

My temperature drop was 20C immediately from 100 to 80c and after time it has lowered a bit further to steadily be around the 70c mark at max load.

My bios was unchanged between the 2 processors. Both ran at 100c during stress testing within moments. The same stress test now peaked at 75c.

No bios adjustments, OC tuning or anything fixed it. The problem was physical.

[LCCF]

then the same as i saw from the bios update. It's still kind of a shock to me i'm seeing temps i never thought i'd see in a SFF case, i'm running low to mid 50C while rendering

be curious if you see any additional improvements if/when you update your bios

[MM12]

Bios was the most recent at the time for both, version 2101 - released Dec 8, 2014

I just checked and there is an updated bios to add support for 5th gen processors, version 2205. I don't think I'll update since I'm happy with the current performance and don't anticipate any benefit.

[GSing15]

So I talked to ASRock about my issue with stock cooling....guess what? They easily replicated it, getting 100 C in 3 minutes. They said they tested an ASUS Maximus VII for reference and got the same results. Why?????