i7 4790K temperatures

[idata]

Hi!

I have a question about the temperatures that I'm getting for my 4790K processor. Like some other people I'm getting extremely hot temperatures when testing the processor at full load.

I am using a Gigabyte GA-Z97X-UD3H motherboard in combination with a Thermalright True Spirit 140 cooler in order to get lower temperatures than the stock Intel cooler.

The temperatures are measured with HWMonitor, RealTemp and the Gigabyte System Information Viewer app. The last one shows slightly lower temperatures then HWMonitor and RealTemp.

The room temperature is around 19°C.

Idle I'm getting temperatures between 20°C and 26°C (depending which core). The Gigabyte app shows even a temperature of 18°C. So far so good.

In what follows I'll continue with the Gigabyte app temperatures, as the graphs are made with the Gigabyte app. Remind however that HWMonitor and RealTemp are showing higher temperatures (around 6°C higher).

As soon as I use Prime95 with the small FFTs test (latest version) to put the cpu under load, the cpu temperature increases instantly to temperatures around 92°C (HWMonitor is even showing 100°C from time to time).

In the graph below you can notice 2 temperature peaks. The first one (lower one) is achieved by running the Prime95 Blend test (max temperature after 3 minutes around 57°C), the second one is achieved by running the Prime95 small FFTs test as described above, with temperatures above 90°C.

The test is done immediately after a cold boot.

The cpu-cooler is at startup running at 550rpm, during load it increases to 1250 rpm.

Another thing I've noticed is that the temperature increases instantly from 18°C to 90+°C, within 2 seconds. As soon as I stop the test the temperature decreases instantly as well: within 1 second the temperature is decreased till 30°C. Next it takes 2 minutes to further decrease to 19°C (idle). This behavior looks very strange to me.

What is going wrong? Why do I get such a high temperatures during load? 90+°C without overclocking is dangerously high. I don't dare to overclock with such a temperatures.

Many thanks for your help!

Edit:

I've done the test with the default BIOS (F4) and the newest BIOS (F7) from Gigabyte.

Edit 2:

In the meantime I was able to do a stress test with the Intel Processor Diagnostic Tool and the Intel Extreme Tuning Tool (10 minutes). Temperatures increased until 71°C, core voltage increased until 1,269V (measured with RealTemp and HWMonitor). These temperatures are better than with Prime95, but still very high.

Edit 3:

After running Intel Extreme Tuning Tool for 1 hour, the max temperature was 73°C. In average the temperature was below 70°C. The idle temperature is around 25°C regarding the Intel tool. Room temperature should be around 19°C.

[BSant8]

Could you please put another link for the thermal test?

The one you posted doesn't seem to work anymore and i deleted the program.

Thank you!

[BSant8]

Hey all!

@karwos - I've rerun the tests, and you are right. The processor stays at 59-60% load while running your test. Please release your 0.2 version and let me know where to download it from.

@lolindir - i'm using the auto bios setting with XMP profile enabled and i've also ran the asus AI Suite3 - 5 way optimisation program. That resulted in an increase of the base clock from 100Mhz to 100,4 Mhz for a total increase from 4400Mhz to 4408Mhz. Also the Vcore voltage offset was set to -0.02 and the OC voltage was set to +0.02. These setting were made by the ASUS AI Suite3 optimisation. Either way, i'm getting the same temps while running Prime95 (ver. 28.5).

[idata]

Becaue i've got confirmation that IntelThermalTest v0.1 was causing only 50-60% of CPU Load, here is new version:

=====================

= 28.10.2014 =

= v0.2 =

=====================

- Load threads have now set proper AffinityMask - CPU Wasn't fully loaded though

- Load threads have now set proper Priority - CPU wasn't fully loaded though

- SW can be less responsive now; after launching test just wait patiently 5 minutes

Download link is same, just re-download and post results : http://www.cyclone-networks.eu/IntelThermalTest.rar http://www.cyclone-networks.eu/IntelThermalTest.rar

Keep noted your PC after launching test may become less responsive (mouse problems, etc). Just ignore it - and patiently wait 5 minutes to get test finished.

[BSant8]

Hey Karwos!

I have rerun the test with the v0.2.

Now the CPU is indeed at 100% load all the time and win7 is almost non-responsive.

I have snapped a print-screen at 97% of the test with the task manager opened and another at the end of the test. Here are the results.

During the test, at 97%:

At the end of the test:

As you can see my temps still don't go over 60. Keep in mind though that i'm using a Corsair H80i closed loop water-cooler. During the test the cooler fans never went above 50%.

I might be a lucky customer that got a good CPU, thought i still don't think it's normal to see above 80degrees C on an unlocked K series CPU while not beeing over-clocked.

Please let me know if you want me to run any more tests and post the results.

Best of luck!

[idata]

What's "normal 100% cpu load"?

100% cpu load means that the cpu is fully stressed. Apparently XTU is not fully stressing the cpu as there is other software which can stress the cpu even more. If some software (like the prime bench) can stress the cpu so much that temperatures are dangerously high, other software can potentially also do this. The only difference is that when using the prime bench you are aware the you are stressing (testing) your cpu and you are monitoring your temperatures. With other, non-bench software, you are mostly not monitoring your cpu and hence you don't know if and when the temperatures are dangerously high.

The fact that there exists software causing such a high load and hence temperatures, means that there might exist more software (now or in the future) causing the same problems.

I've also read the stories about the bad connection between the cpu itself and the heat spreader. I would like to verify this but the only way how to verify this is by violating the warranty policy and risking a destroyed cpu.

I am considering to RMA my cpu but since it seems to be a structural problem I doubt this will solve anything. I could of course keep RMA'ing my cpu until I have a good one, but than I might end up for a long time without cpu which is a problem as I need this pc for our business. I'm also living abroad so it's not easy for me to reach my supplier where I've bought the cpu. I don't know if you can contact Intel directly for an RMA?

[idata]

I didn't have time yet to test my cpu, as I only have an Ubuntu installation at this time and I am using my system as a production system. I hope I'll find some time this weekend to install Windows next to it and run your test.

You don't have a Linux version of your test suite by accident?...

By the way, notice the voltage differences from Brake and Woolphy. Are you guys using stock settings in BIOS?

[SPark48]

This is my first on the forums and would like to add my experience with this issue.

I have been having the same problems with overheating i7 4790k with an Asus Z97 Pro motherboard. The idle temperatures are in the 30C's and ambient temps are around 22C. There are 4 120mm Corsair Performance intake fans and 2 80mm exhaust fans in the case to also help keep things cool. First, I started off using the stock cooling fan and at stock settings without ANY overclock, the CPU was hitting high 90C's and up to 100C running CPU stress tests within the first few seconds of running them before I quickly shut it off. These tests include Prime95, OCCT, EVGA OCX Scanner, and HeavyLoad. Every single one of those stress tests gave the same results. After checking to make sure cooler was on correctly and reapplying paste, I checked again and there were no changes. So, I swapped out the stock cooler with a SilverStone NT01-Pro with 4 80mm cooling fans attached and still, the temps were in the 90C's after 3-4 minutes of running said tests. I then went out and purchased a Corsair H60i liquid cooler and now the temperatures are hitting 89C-92C. I've taken out and reseated all the coolers using different types of "top-of-the-line" pastes, yet nothing changes the temperatures from reaching in the 90's.

So, last week I contacted Intel customer support. They sent me an email asking me to download and run the Intel Processor Diagnostic Tool. The test ran and passed with temps reaching mid 60C's. When I responded back to the email with results of tests along with pictures of temps hitting 89C with OCCT, the response was "[T]he processor is fine and the temperature with other tool shows that the processor is high but in this case if the Intel® Processor Diagnostic Tool passed it means that the processor is fine and the temperature is below maximum". That's the direct quote from the email.

Am I supposed to be satisfied with this response? Does anyone have any experience with similar situation? I'm not sure what the Intel PDT tests, but I highly doubt it is taking the CPU to max or even high CPU loads. Intel has publicly stated that the i7 4790k should have no problems overclocking to 5.0Ghz, yet we can't get the temperature to stay cool on STOCK, yet "the processor is fine"?? How does that make any sense? I have responded to the above email to let them know that I am not satisfied with that response. I will update on how they respond.

[BSant8]

Hey Sampark and welcome to the forums.

I'll post my ideas about your setup maybe some might help.

1st of all, i would have more fans getting hot air out of the case then fans blowing air into the case. From what i read, a slightly negative pressure in the case is better for overall temperature of the system.

2nd you should put some washers between the motherboard and the backplate of the cooler. Some peope reported that the mounting of some corsair water cooler on some z97 boards are not perfect. I managed to drop my temps by 10-15 degrees after i installed the washers.

3rd keep in mind that some thermal compounds have a break in period. If that is the case and the cooler is mounted properly, you should see some slight drop in temp after a while. Not that much though, between 2 and 4 degrees.

Next, run the intel test again an check the workload in performance tab (ctrl+shift+esc). It is true that the new stress tests use some AVX instructions which generate A LOT of heat but that still shouldn't be a reason to overheat and throttle.

Also, try the program that kawos created and post the results here.

Best of luck to you!

[SPark48]

Woolphy,

Thank you for the response.

1) I will try changing the direction of a couple of fans tonight, but in my experience with computer builds, changing the direction of intake and exhaust fans have a negligible affect on the CPU temperatures. We're talking 1-2C differences sometimes. I am seeing significant high temperatures of up to 100C.

2) I have heard of the issue with Corsair water coolers and I have tried the recommended ideas such as tightening or loosening the screws on the backplate. I haven't tried using a washer yet, but at this point, I'm willing to try anything. My concern is that I'm seeing these temperatures with the stock cooler as well as the SilverStone NT01-P. Also, I have a computer running an i5 4690k and it has no problems at all with the same H60i liquid cooler.

3) I am aware of the break in period of compounds. As you stated, it should not have my temperatures going in high 80C's-90C's due to this.

4) When running the Intel PDT(64bit), the performance tab shows it running 92-94% of the CPU with other programs running in background adding to about 98% load. How is it that the Intel PDT is so significantly different than the other CPU stress tests out there (Prime95, OCCT, EVGA OCX, etc.)? Also, in real-world testing, games such as BioShock, Ghost 2, even NBA2k15 gives temperatures of 80C+. Is it possible that the Intel PDT is running single core test rather than running all 4 cores?

Thanks for your time.

[SPark48]

As you suggested, I ran the Intel Thermal Test. The first time it ran, it ran fine and passed with max temperature of 66C. However, now it is crashing my computer every time I run it... Here are the results from the first test that passed:

Does this mean that my processor is fine and there is nothing wrong with it? My concern is that 1) Why is the test crashing my system? and 2) Why are the results so vastly different from other CPU stress tests such as Prime95, OCCT, EVGA OCX, HeavyLoad, etc.? Here are my results from OCCT that I ran right after the Intel Thermal Test with absolutely no setting changes:

Notice how after it hit 92C it throttled the CPU and went back down to 70C's. I cancelled the test soon after the throttle just to make sure no further damage is done to the CPU. Why are these two results, ran right after the other, show such vastly different results? Should I be satisfied with these results?

[BSant8]

Hey Sampark!

No, i don't think you should be satisfied with the results. 90+ deg is more than allowed for a CPU. The difference between Intel Thermal Test made by kawos and the rest of the stress test programs is that the rest use those AVX instructions which are extremely stressful on the CPU. Another thing is that the data that is used is quite small and fits in the L3 cache of you processor so the processor does not have any breathing period and works at 100% load all the time. I would be curious to see if all of the prime95 (ver. 28.5) stress tests generate the same heat. Could you try to run the Small FFT test, Blend test and Large FFT test and see the difference in temp? You should also check the vCore voltage while running the tests just to see how much current is given to the CPU by the motherboard.

Also, until someone can find a viable fix for this, i would set the vCore offset voltage to -0.02 or -0.03. That is what i did after i saw that there was no fix yet. Your ASUS motherboard should have this option somewhere (UEFI, BIOS, Ai Suite 3). By doing this i managed to keep my CPU running stable at 4.4ghz with temps of 80deg max on Small FFT test. Sometimes the motherboards feed too much voltage to the CPU even though it does not require it, thus generating more heat. The hotter the CPU runs, the more amps/volts it needs to keep the speed, thus generating even more heat (at least this is what i found online while reading about this issue).

Hope it helps!

[SPark48]

You have been very helpful, and I really do appreciate your time.

Do you know what programs use AVX instructions? If none of the programs I regularly run use AVX instructions, I should be OK, right?

I will run your suggested Prime95 tests tomorrow and post the results here. As for your suggestion of decreasing vCore voltage, I followed the same suggestions from other forums and did that for now to make it more stable and "safer" for the CPU. However, I honestly don't think that should be the answer Intel gives to solve the issue; "Just decrease your voltages and you'll be fine". They really have to be embarrassed at how this has turned out, and if they're not, they should be. I've been able to overclock the Pentium G3258 to 4.5GHz and the i5 4690k to 4.7GHz with no issues. Yet here we are talking about struggles with out-of-the-box temperatures on one of Intel's higher-end unlocked processor here - the flagship of their 4th generation consumer processor family. Intel has said this processor can be overclocked to 5.0GHz, but I haven't even had the courage to try overclocking this yet due to these temperatures on stock settings. If we can't even overclock this processor at safe temperatures, what's the point of purchasing a K CPU? I may as well have paid less and purchased a regular 4790.

Judging by the sheer number of people that are having this same issue with the same processor, I don't think they can turn a blind eye to this. Eventually, they will have to address it. I just really hope they don't do what they recently did with the Pentium processor issue, which is pay back a measly $15 and tell us to be happy with that...

I'm still waiting on response from Intel customer service. Will update when I get a response.

[SPark48]

Speaking of the Pentium G3258 and i5 4690k, why is it that these two processors have no issues with AVX instructions, yet the i7 4790k struggles so much with them? I have run the same tests on those two other processors and have had no issues with temperatures averaging in the mid 60C's. That's why I think it is a manufacturing issue of some kind with the i7 4790k's and not just simply that AVX instructions are stressful on processors. If the issue was AVX instruction stress, in theory it should give the same high temperatures on the other processors as well, right? I have literally put both of those other chips - G3258 & 4690k - into the same exact system (same case, fans, cooler, everything) and there was no issue. Yet, put the 4790k in and temps are back in the 90C's. Their "the processor is fine" response just doesn't add up.

[idata]

I had a bit of a break in my answers, cause I was waiting for feedback from Intel support and installing Windows 7 again on my system.

I've just got a nice call from Intel support and basically this is what the man told me (summary):

• The temperatures read by the software might be faulty. You cannot trust them.
  • I've asked the man if this means that also the official Gigabyte tools + BIOS) are reading the temperatures wrongly, but here I only got the answer that I should verify this with Gigabyte. However both (third party software and Gigabyte) are using the same sensors on the CPU I suppose?...
• There is not specified how warm the cores may be, only TJMax and Tcase. Since I cannot measure these temperatures I cannot say whether or not the CPU is working as specified.
• As long as the CPU is not slowing down or shutting down the system, nothing is wrong and the temperatures are normal. If the temperatures would be too high the CPU will automatically shutdown or throttle.
• A K-processor only means that you are able to overclock, but it doesn't mean you can overclock (a lot). This depends from CPU till CPU. I might not be lucky with my CPU. He advised me to buy an extra overclock warranty from Intel in case I would like to overclock.
• Since my cooler is not feeling very hot, it means that the CPU temperature should be fine.
  • My thought: or that the cores and the heat spreader are not having a proper contact, and hence not removing the heat efficiently.
  • Can some of you feel at your CPU cooler if it becomes very hot in case you are running Prime? Very hot = I can actually not touch it without almost burning my hand.
• If the Intel processor diagnostic tool is running fine, there is no problem.
  • Has someone tried to run this tool in combination with Prime95 (simultaneously)? Cause you should in this case get extreme hot temperatures. I wonder if you than still get a "passed" message from this tool...

Conclusion:

My CPU is fine, temperatures are normal and nothing to worry.

At least, regarding the Intel guy...

[LKocz]

Hello. I'm adding myself to the list with the same problem. Just got it last night. Temperatures 70C and above under heavy load on stock cooler. Didn't get chance to test it further yet.

[Allan_J_Intel1]

To all,

Check Ken notes posted on Nov 3, 2014.

/message/260210# 260210 https://communities.intel.com/message/260210# 260210

Allan.

[BSant8]

Hey Sampark!

I'm sorry i didn't answer sooner but i was busy with work. I saw that you got your CPU RMAed. I hope you get a better one now.

On the subject, the main problem this CPU has is it's heat-spreader. Since with this CPU the chip is not soldered with the heat-spreader, intel used a thermal compound which is between the two. I'm guessing the space between the chip and the heat-spreader is a little to large for an optimal transfer. Unfortunately, if that is the case, there's nothing we can do about that for now, unless we want to void our warranty and delid the CPU.

On a happier note, i had a little bit of time today to play with the settings on my CPU and managed to get a 4.8ghz speed on it with a vcore of 1.28V. It was stable while running the intel XTU stress test for one hour (temps in the mid 70s) but when i tried Prime95 it crashed after 15 min :-). So there's something strange about how Prime95 stresses the CPU or how Intel XTU stresses the CPU. One of them does it wrong and i can't figure out which one.

Best of luck :-)

[SPark48]

Hey Woolphy! I am following you on the other thread as well. It seems busier than this one so we should probably take the discussion there.

As far as my situation, yes I threw a fit and they begrudgingly agreed to a "courtesy" RMA exception. I also hope the new one gives better results.

I am also concerned about the 3rd party stress tests such as Prime95 and OCCT. Both of these tests are either crashing or giving extremely high temperatures (90C-100C). I have not overclocked the processor yet as I don't want them to charge me for another processor when they get this one back, but I am hopeful that the suggestions posted by Ken on the other thread are showing some improvements.

If it is a hardware issue with the heat-spreader, then I believe Intel needs to address this like they just did with their Pentium processors. I just hope it's more than the USD$15 they paid 10+ years after the problem was first noted. I'll keep you posted on how things go with the RMA. Thanks again!

[LSala4]

so is 70 degrees celsius good for full load? I currently get 22-27 degrees celsius when idle and a constant 60-70 using prime95 for 20 mins, idk if thats good since in your last edit, you claimed 25 celsius which im not sure if the edit was to say its fixed or if your still having problems

[idata]

Sounds good to me, if you're getting 60°C-70°C after 20 min running Prime95 with small FFT's.

My processor is in RMA at the moment because I have spontaneous reboots after encoding video with Handbrake after approx. 1,5 hours of encoding (Ubuntu 14.04). Hopefully these reboots will be solved, as well as the hight temperatures.

[idata · Accepted solution]

Hi all,

Since yesterday I have a new CPU (RMA). I've done some (short) tests with Prime95 small FFT's and I can confirm that based on these tests, the max temperature seems to be 10°C - 15°C lower. Without any BIOS adjustments I'm getting temperatures beginning 80°C. I have to admit that besides the new CPU I'm also using other thermal paste (Arctic Silver 5), but this shouldn't change the temperatures a lot: maybe 1°C or 2°C.

Next I've applied the recommended Intel settings from Ken (Intel support), which can be found back in the topic mentioned before (/message/260210# 260210 https://communities.intel.com/message/260210# 260210). You can find a summary of the proposed settings in this official Intel document:

2 changes I've made on my Gigabyte UD3H mobo are:

• I've put the VCORE voltage on 'normal' instead of manually adjusting it (because otherwise the voltage remains all the time the same instead of dropping when idle). Instead I've used a negative offset of -0,060V. This seems to run stable, however I didn't do long duration tests yet. But I didn't have any system crashes during testing yesterday evening, nor in idle, nor in stress. It might be even possible to decrease the negative offset even further (-0,070V or maybe even -0,080V).
• I've activated the XMP1 profile of my RAM. This could increase the power usage a bit (and hence also the temperatures), but since I am using Crucial Ballistix DDR3L memory running @ 1,35V (even when activating the XMP1 profile!) I don't expect any (significant) negative effect.

After applying these changes I've notice that the CPU speed is limited to something between 4,10 and 4,20 Ghz instead of the default 4,4Ghz (with default BIOS settings). These new values are actually correct, as 4,4 Ghz should, regarding Intel specs, only be achieved when only 1 or 2 cores ares stressed, not when all 4 cores are stressed as with Prime95 Small FFT's. (I didn't do any single core stress test yet.)

The temperatures I'm getting with these settings are even better. After running Prime95 with small FFT's for a couple of minutes, I'm getting stable temperatures around 64°C!

Notice this is achieved with an aftermarket air cooler, not with the stock cooler. I am using a Thermalright True Spirit 140 BW cooler.

I will do some more testing and if everything seems to be stable, I'll forward my settings to Ken (Intel support) so they can add them to the official documentation mentioned here above. I'll also mark the link to this thread as 'correct answer', but keep in mind that it's a combination of a new CPU with this settings which solved my problems. Without the new CPU I would expect that the temperatures would be 10°C - 20°C higher.