Is my i7-7700K overheating?

DukeLaCroix

Prominent
May 27, 2017
7
0
510
Hi guys,

two days ago I installed my new i7-7700k cpu and I just had time to do some stress tests with it. I ran a game as well as 3D Marks CPU benchmark and prime95 v26.6 torture test.

During 3D marks benchmark, CPU hit 99-100% usage and got as hot as 88° celsius.
During prime95, it got as hot as 93-94°C.
During the game I could test (playerunknown's battlegrounds), it was mostly at 30-50% usage and got around 75-82° celsius.
Ambient temperature is around 22°C and idle temperature around 45°C.

I have a Scythe Mugen 3 rev.B mounted on the CPU in a be quiet! Pure Base 600 case which has four be quiet! PURE wings 2 fans. 1x 120mm & 1x 140mm intake (both front) and 2x 120mm exhaust (back and top). I know that the CPU fan is mounted properly and fans were spinning at max RPM (1600 for the CPU and 1200 for the case fans).

My question:
Is the CPU overheating and in how far is the setup fit for a potential overclock to, say, 4.5 Ghz?
I checked back with the Intel tech who told me that the CPU is design to run up to 100° celsius so I should worry. I wonder though in how far I will remain under 100° celsius with this setup once I want to overclock.

I presume part of the answer is that my case is not optimal for cooling and maybe that my CPU cooler is not the best on the market, but other than that: should I be worried ?
 
Solution
7-8c sounds pretty significant in terms of side panel on vs off. Maybe the case fans aren't moving air through the case as well as they could be. 1.27v isn't abnormally high, it could be higher if overclocking though it sounds a bit high for stock. If manually setting it in bios try setting it around 1.2v and run your stress tests. Maybe try rog realbench, it's a fairly good stability test that runs through multiple scenarios. If it's stable, try 1.18v and so on. If it blue screens or fails a test then slightly raise the voltage up a little on the next reboot.

Some do tend to run hot but delidding can be a somewhat tedious adventure and will void the warranty. I'd maybe consider reading further on that and opt to delid more as a last...

genthug

Honorable
Stress tests on a non overclocked CPU should not have you hitting 90+ C. I would not overclock that unless you can achieve the overclock without changing the stock voltage settings and that is after you're able to bring that temperature down into the mid to lower 80s to be on the safe side. Are you sure you have the cooler mounted correctly and that you correctly applied the thermal paste?
 
That's too hot and overclocking is out of the question if you're already so close to 90c+ as it is. What are your ambient room temps, are they on the hotter side (around 30c or hotter)? Preferably the cpu at stock wouldn't exceed 75c much less 82c while gaming at 100% use, you're already hitting that at 50% use.

Technically the cpu will start to thermal throttle at around 100c which means automatically slow down and underperform to get temperatures back in the safe range but that's more or less a safety feature, not a goal. The hotter a cpu runs like that all the time over the years it will likely reduce its lifespan and cause it to fail sooner.

You might try checking the core voltage and manually setting it in the bios, maybe the bios is auto selecting core voltage that's too high for the speeds you're at and causing excessive heat. Check in the bios for cpu fan speed settings and see if there's a way to allow the fan to spin faster.

It might be worth making sure you've got some thermal paste on hand, remove the cpu cooler and clean both the cpu's metal ihs lid and the base of the cooler with a lint free cloth and isopropyl alcohol, then reapply thermal paste using a small pea size blob in the center. Try to make sure the retention screws for the cooler are tightened a little at a time back and forth for even mounting to the cpu.

Have you tried checking the temps using something like realtemp and are all the cores the same or is one core considerably (10c or more) hotter than the others?
 

DukeLaCroix

Prominent
May 27, 2017
7
0
510
Hey all,

thanks for the super quick replies already.

@Genthug & synphul: I am pretty sure the CPU is correctly mounted (the screws are rock solid) and I applied the thermal paste correctly (small pea size blob in the center).
@synphul: I also doublechecked temperature with HWMonitor and compared it to Core Temp. They were both in line and although one core was usually a few degrees higher than the other, it never exceeded 10°C. As written earlier, the ambient temperature is around 22°C, given or take 1-2°C (I unfortunately have not a very good thermometer).

The BIOS is also flashed to the most recent version and voltage settings are running on default (AUTO). I might indeed play around with that a bit, but I am certainly no expert and don't want to break anything here.
I can indeed re-setup the whole CPU&CPU Cooler on the weekend to be absolutely sure everything has been setup properly.

Would a different CPU cooler maybe better (generally, not necessarily for OCing)?
 

Rakanyshu

Distinguished


Even with bios update i find that the motherboard still puts more voltage than the cpu really needs for stock or even oc performance, once you start to oc the cpu and manually tweek vcore you will find it running at a lower temp. in my case the gigabyte aorus gaming k5 mobo came with 1.4 vcore at stock, after bios update it lowered to 1.295 and was still runing a bit hot so i manually lowered to 1.205 and it stays stable at stock speed and turbo bosst on.
 

DukeLaCroix

Prominent
May 27, 2017
7
0
510
Ok, so I re-mounted the CPU and CPU cooler to doublecheck and, unfortunately, it didnt do anything. Temperatures are still the same.

I checked the VCORE settings and the voltage under full load never exceeds 1.270 V. That doesnt strike me as abnormally high...

I noticed though, that the influence of the be quiet! PURE base 600 case can make quite a difference because the temps were around 7-8°C lower when the two side panels were off the case.

Nonetheless, even around 80°C+ when gaming and around 86°C at full load prime95 stress test seem higher than I would have anticipated...

Do you guys maybe have any other ideas what might cause this?
 
7-8c sounds pretty significant in terms of side panel on vs off. Maybe the case fans aren't moving air through the case as well as they could be. 1.27v isn't abnormally high, it could be higher if overclocking though it sounds a bit high for stock. If manually setting it in bios try setting it around 1.2v and run your stress tests. Maybe try rog realbench, it's a fairly good stability test that runs through multiple scenarios. If it's stable, try 1.18v and so on. If it blue screens or fails a test then slightly raise the voltage up a little on the next reboot.

Some do tend to run hot but delidding can be a somewhat tedious adventure and will void the warranty. I'd maybe consider reading further on that and opt to delid more as a last resort. With case temps being a bit high you might need a better cpu cooler. Some cases just run hotter than others by design, especially if they're designed to be 'quiet' cases. They usually take an approach to limit noise escaping the case which hampers airflow/cooling a little vs a more open design.

It was reaching 80c+ when gaming even with the side panels removed? The mugen 3b isn't the worst cooler out there but there are others which perform better. In some benchmarks it cooled only slightly better than a 212 evo. In a quieter case trading cooling for noise reduction it could just be a combination of things. Less than optimal case airflow, average cooler performance and a cpu which may be running hotter than others. A variety of reasons all creating a situation where a couple degrees here and there are adding up to cause it to run 10-15c hotter than it should be.
 
Solution