Intel Core i7 4790K reaching 90C, When overclocks. Need Help!!!

[OneMoar]

hes from Pakistan as well ambient temps are what 33 or 35c

a Noctua NH-U14S is not going to cool a 4790k @ 44x 1.250v it just isn't

the temp variations are probably thermal or power throttling

op needs a better cooler and his chip looks like a candidate for de-lidding and clu
./thread

 

[Kursah]

4790K @ 44X multi with 1.250v is stock for quite a few of them. 44X is the max multi for single core though, and obviously does the Intel trickle down as more cores are pushing turbo clocks at that point. Syncing all of them to 44X will increase the load, but that U14S should handle it. It has in all the situations I've deployed them in without issue, mine being the only one I added a second fan to and really it made little difference. Hot ambient temps though, that is a game changer and a higher capacity or type of cooling might be required to battle that successfully.

Really no reason a U14S shouldn't be maintaining decent temps on that CPU running so close to stock specs, now the stock cooler I can see struggling and reaching the thermal limit quite easily and witnessed that dozens of times over. A 100MHz OC at near stock load voltages shouldn't be killing this setup quite this badly, unless ambient really is that hot and airflow is really poor and the CPU fan is on the lowest RPM available in firmware with no ramp up or not plugged in lol.

Thermal throttling won't hit until 100C on that CPU, at least from my experience with the Haswell and Devils Canyon CPU's that's how they've worked. Once thermal throttle is hit, it'll drop in 100Mhz increments until it can maintain just under the thermal limit. Again, that's been my experience with these series of Intel CPU's.

 

[BiggieShady]

Please help me identifying the issue, whether it is mobo or cpu, creating problem.


Regards,
The N

It's not the motherboard, that's for sure ... and your temps seem in normal range for that chip/voltage/multiplier/cooler combination in prime95
According to you, you are having plus 15 C temperature from what you had back then on same settings.
Can you verify that there is no difference in ambient temperature in those measurements? Difference in dust buildup?

 

[OneMoar]

depending on the boards default current and power limits they can throttle at any temperature out of the box my board will pull power at 4.2ghz @ 75c because the current limit is set low for my chip-bin

and remember turbo speeds are NOT base frequency
https://www.intel.com/content/www/us/en/support/articles/000005523/processors.html

4/8t @ 4.4 is a lot more current and heat then stock turbo, there is also the chip bin to consider he looks like hes got a mediocre chip

his options are to de-lid, install clu and relid it (there could be something wrong with the tim under the ihs even on the **90k's its not exactly great )
or MOAR cooling

the later is safer if he doesn't have access to proper delidding tools

 

[EarthDog]

That cooler shouldnt have an issue at 1.25v 4.4ghz. Its not a damn potato. In fact, its 3C off the NHD-15 with a 150W load....
https://www.anandtech.com/show/11407/140-mm-slim-tower-cpu-cooler-roundup/8

Though he is in Pakistan and there are warmer temps, we dont know if the dude has AC. Likely considering the hardware. OP what are your room temps?



Edit: just noticed.... the difference in screenshots in the first post... one seems to be using 8 threads, the second, cooler one, is using 5(?) threads.... thats the difference (assuming im seeing things right).

EDIt2: 4 threads? Not sure.. window looks smaller too so perhaps there are more under? I don;t know. Also, using Realtemp in one shot and Hwmonitor or w/e in another... that could also be a part of it.

 

[cucker tarlson]

30-35 ambient can cause a yuge temperature rise compared to 20-24 most of us have , close to 20 degrees on cores. 4790Ks are hot as hell, especially in stress testing in such conditions on a decent but definitely not top of the line cooler.

 

[Gasaraki]

Hey Guys,

I have been observing some abnormal behavior on i7 4790K prcessor. It was all good a couple of days ago, but since yesterday, this processor producing an abnormal behavior with no clue why.

Let me get straight to the subject. When I overclock i7 4790K at 4.5GHz with 1.12vcore in bios, and put it on torture test of Prime 95 25.11v, it jump straight to 85C-90C within 3 minutes. On the other hand, it touches 70C at stock clocks. which also seems a bit higher to the previous 62C when it was behaving normal. I asked a buddy, he says, Chip problem. But I am sure if that's completely right.

Also, I tried to reseated the Cooler, Noctua NH-U14S, a couple of time, and re-apply thermal paste twice with dot and spread technique. Even, I tried a different Air cooler, too, but nothing has been changed.

Some screenshot of load temperatures:
i7 4790K @ 4.5GHz
View attachment 93022

i7 4790K @ 4.0GHz (stock)
View attachment 93023

My BIOS settings for overclocking are:
Multiplier x45 as Sync All Cores
Voltage 1.25v
DRAM 2400mhz (default voltage)
Turbo enabled (It get auto enable when increase the multiplier from stock)
EIST disabled (by default it is Enable)

All other options have been untouched.

Previously, when processor was doing normal, I overclocked i7 4790K at 4.5GHz and settings used were same as above. But now, it producing issue, no idea, why.

Here is teh temp screenshot for past overclocking being maxed out to 75C. (same system config)
View attachment 93024

System Config:
ASUS Maximus VI HERO
Core i7 4790K 4.0GHz (subject)
Gskill 8GB 2400mhz
Enermax REvolution XT 650W GOLd
Noctua NH-U14S Cooler
Noctua Nt-H1 Thermal Compound

Please help me identifying the issue, whether it is mobo or cpu, creating problem.


Regards,
The N

So you are getting mid to high 70s @ 4.0 but getting high 80s (on one core) @ 4.5. Not seeing a issue here, try push pull on that cooler to get better cooling.

I don't know why you keep saying 4.5 overclock when in your screenshots it's different. The low temp screenshots are running at 4.0 (with the multiplier at 40x) and the high temp ones are at 4.5 (45x).

 

[EarthDog]

30-35 ambient can cause a yuge temperature rise compared to 20-24 most of us have , close to 20 degrees on cores. 4790Ks are hot as hell, especially in stress testing in such conditions on a decent but definitely not top of the line cooler.

Actually, its close to 1:1. If there is a 10C increase in ambient, your temps should go up ~10C. Science. Its why many review sites are able to 'normalize' their temperature to a given value.

 

[John Naylor]

I agree with the above poster ... i recommend strongly against using synthetics for overclocking.

1. What's the point ? If you built ya PC to get ya name on web site leader boards, many of these still require validation by synthetics. so no option. But if you built ya PC to run applications / games, then and application bases stress test will be more useful and present more realistic data with regard to what your PC will see in its lifetime.

2. I have had 24 hour P95 stable OCs fail in 45 minutes under RoG Real Bench Multitasking test

3. IIRC, P95 ver 26.6 was the last version w/o AVX. Using newer versions is risky as it's capable of damaging CPU if not careful. So if ya use P95 w/o AVX and pass, you have proved that your CPU is stable .... but only as long as your programs don't call for use of any modern instruction sets. So what happens when those instruction sets are called for ? RoG RB includes these instruction sets using ral applications in a mutitasking environment/ So while it will stress the CPU in ways it will likely never see again, it will produce core temps 5C or mor elower than what synthetics are capable of.

Your cooler is a pretty decent performer and should easily outperform 120.140mm CLCs. If your case allows, ya might want to try a Scythe Mugen Max or Fuma which Top the NH-D15 and most 240mm CLCs.

Your temp differential is more than can be accounted for by TIM. Application techniques are not universal and should be tailored to the type of heat sink
http://www.overclock.net/t/1621347/kaby-lake-overclocking-guide-with-statistics

How is case cooling ?... one way to test this is to take side panel off and a decent desk fan (i.e. Vornado) and retest with the fan on max ... if temps change, more case cooling is recommended.

 

[The N]

Hello, Thank you guys for feedbacks. Well, I have been checking my CPU for past 2 days, on P95 and AIDA loads. tried a different heatsink cooler; SilverStone Heligon dual tower. I think my chip can't afford to cool down from those Celsius, while at overclocking temp shoots up to 85c and touches 90c and still going.

In game and normal loads, temps are very normal. But nothing has been change in stress testing.
The only solution I could see is to avoid using Synthetics stress test.

This haswell is way out of control at least in my case. I might upgrade this to much cooler chip.
Do you guys think I should upgrade to Skylake, now ?

 

[EarthDog]

Skylake isnt going to be any different.

 

[FYFI13]

On top of that, performance increase would be neglible. If you really want to upgrade get I7 8700K and do NOT run synthetic bencharks, for peace of your own mind

 

[EarthDog]

In game and normal loads, temps are very normal. But nothing has been change in stress testing.

Just caught this...

Things are NORMAL. Stress testing will show higher temps than gaming. Its NORMAL.

 

[John Naylor]

Hello, Thank you guys for feedbacks. Well, I have been checking my CPU for past 2 days, on P95 and AIDA loads. tried a different heatsink cooler; SilverStone Heligon dual tower. I think my chip can't afford to cool down from those Celsius, while at overclocking temp shoots up to 85c and touches 90c and still going.

In game and normal loads, temps are very normal. But nothing has been change in stress testing.
The only solution I could see is to avoid using Synthetics stress test.

This haswell is way out of control at least in my case. I might upgrade this to much cooler chip.
Do you guys think I should upgrade to Skylake, now ?

I don't see the concern. If a soccer mom is looking for a van to bring the kids to practice in sunny Florida, is there any value to her 'stress testing" it off road on 4 wheel drive tracks and in heavy snow ? It's not about changing CPUs, it's about using realistic testing criteria applicable to your situation. If you open a delicatessen and sell cold cuts, you don't need to test ya scale with 50 pound loads ... no one is going to walk in and buy 50 pounds of salami.

a) Synthetic stress tests place loads on your CPU that it will **never** see again in its lifetime
b) You can be 24 hour P95 stable and still fail in a realistic multitasking application based stress test like RoG Real Bench
c) Synthetic stress test can produce temps 10C or more higher than an application based stress test. So if say you don't want to expose your CPU t anything above 80C, you would scale back your your OC if you saw 85C under P95... but under RoG Real Bench, you'd likely see something in the mid 70s ... in gaming, you'd be in the mid 60s.

 

[eidairaman1]

Just Use Ryzen Blender or OCCT, if it passes those without lockups or bsods you are fine.