• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

40th Anniversary Intel - Core i7-8086K - 5.0GHz+

Rroff said:
There have been plenty of people over the years complaining after 1-2 years of running high voltages of BSODs and having to turn down their overclock to get stability, etc.

Heat also tends to be bad for EM.

EDIT: The datasheet for this processor does note 1.52V max with the notation "Long term reliability cannot be assured in conditions above or below Max/Min functional limits.".
Click to expand...

yep its voltage degradation. Motherboard manufacturers wont care as they just need to make it to the year warranty. Plus most chips can degrade and still meet stock speeds.
 
Is it worth delidding?

Currently at 5.1 Ghz VID showing 1.36V in HWINFO and 1.344V in CPU-Z. Temps about 80 using IntelBurnTest and stable.

Not sure whether to add more voltage to make 5.2 stable (needs a big jump I think, closer to 1.4V in the BIOS), or take it down to 5 Ghz and lower Voltage to the point it becomes unstable and then put it up a little.
 
Just get the delid mate, thermal grizley liquid metal and a tube of heat resistant silicon. Use nail polish or liquid electrical tape to cover anything that could short under the IHS. Theres loads of videos showing how to do the 8700k. When you reattch the IHS just put a tiny bit of silicon on the 4 corners so its easy to remove next time if you need to.

My chip isnt the best but can hit 5.2ghz at 1.45.
I just leave it at 5ghz 1.39v for every day use.
 
Cheers, I have ordered all the bits.

I've dropped the clock speed to 5 Ghz and started lowering the voltage. It's now stable with VCore set to 1.36V. Using Intel Burn Test on very high it passes without issue, voltage in CPU-Z is showing at 1.3V and VR OUT in HWINFO records a 1.354V maximum. Happy with that, think it needed 1.38 something in the BIOS for 5.1 and I couldn't get it stable at 5.2 Ghz even up at 1.4V. Temps about 80, 88 max on one of the cores. I'm hoping the delid will drop it into the 60's if not I fear my CPU block might need some maintenance. That said, it's currently stress testing and the CPU temperature is 66, so I guess it spikes a little.
 
Well it's all done now. Quite a drop in temps! Might crank up the clockspeed and voltage and see how far I can get.

image.png
 
Hey guys :)

Can someone give me a hint here?
I've been testing the lowest voltage I can get my 8086k to run with 5Ghz (4.7Ghz cache) on all cores and the best I could do was 1.25v llc2 (asrock) which ends up at 1.248v when idling and 1.216v at full load.

I'm super happy with the results and with the chip itself but I've noticed that right after the end of RealBench stress test my idle voltage jumps to 1.268v for less than a second and then goes back to 1.248v. This only happens after the stress test is complete. If I use the PC normally for gaming and so on the vcore never goes above 1.248v but on the days I use RealBench for test stability it will reach 1.264v very very very briefly after using it.

Any obvious explanation for this?

Thank you in advance.



PS:. I think i have a very good chip. I can run cinebench and game at 1.168v llc1 but to be realbench stable the best I can do is 1.25v llc2 (1.216v full load)
 
I've droped vcore to 1.245v and it never reaches 1.268v now.

Other thing i've notice it that while running realbench the vcore sits on 1.216v but if i run IntelBurnTest or prime 26.6 FFT 1344 it sits on 1.232v
This means that realbench produces more load than the other tools?

 
nunocaires said:
Hey guys :)

Can someone give me a hint here?
I've been testing the lowest voltage I can get my 8086k to run with 5Ghz (4.7Ghz cache) on all cores and the best I could do was 1.25v llc2 (asrock) which ends up at 1.248v when idling and 1.216v at full load.

I'm super happy with the results and with the chip itself but I've noticed that right after the end of RealBench stress test my idle voltage jumps to 1.268v for less than a second and then goes back to 1.248v. This only happens after the stress test is complete. If I use the PC normally for gaming and so on the vcore never goes above 1.248v but on the days I use RealBench for test stability it will reach 1.264v very very very briefly after using it.

Any obvious explanation for this?

Thank you in advance.



PS:. I think i have a very good chip. I can run cinebench and game at 1.168v llc1 but to be realbench stable the best I can do is 1.25v llc2 (1.216v full load)
Click to expand...

Thats golden sample territory. I will take your word for it is proper stable and assume you havent cheated with AVX offsets.

If you want a comparison you are stable at 1.21v 5ghz, vs my 1.36v 4.8ghz. Intel have way too much variability on their chips, that gap is a chasm.
 
chrcoluk said:
Thats golden sample territory. I will take your word for it is proper stable and assume you havent cheated with AVX offsets.

If you want a comparison you are stable at 1.21v 5ghz, vs my 1.36v 4.8ghz. Intel have way too much variability on their chips, that gap is a chasm.
Click to expand...

Hey friend. What do you mean by cheating? I might be unaware of what you mean. With all my honesty, I'm running it with AVX offset of 3. I only did lower it based on what people usually suggest online. Should I try without the offset?

Regarding stability, At the moment, at 1.245v llc2 (1.16v-1.32v on full load depending on the software) I can do 10 runs of IBT (Maximum) and 8h RealBench.
 
chrcoluk said:
If you want a comparison you are stable at 1.21v 5ghz, vs my 1.36v 4.8ghz. Intel have way too much variability on their chips, that gap is a chasm.
Click to expand...

If it makes you feel any better mine will only do 4.8GHz at 1.26v with an AVX offset of 2. Without it I have to do 1.3+ and that's after a delid, it really needs a lot of power. There is indeed huge variability.
 
My poor 8700k can spike to 90c
It’s 5ghz 1.37v on a 360mm aio

But that’s synthetics - games is reasonable 45 to 65c

I can’t justify delidding as I won’t pish any higher as the voltages are already high
 
nunocaires said:
Hey friend. What do you mean by cheating? I might be unaware of what you mean. With all my honesty, I'm running it with AVX offset of 3. I only did lower it based on what people usually suggest online. Should I try without the offset?

Regarding stability, At the moment, at 1.245v llc2 (1.16v-1.32v on full load depending on the software) I can do 10 runs of IBT (Maximum) and 8h RealBench.
Click to expand...

Ok thanks for the honesty.

You do what you want to do its your system.

Cheating is perhaps an overly strong word and I will take that back, rather I feel its a bit misleading if you quote a speed without saying you using an offset, but I feel so much software now days uses AVX, then effectively your true overclock is the whatever the AVX offset is, e.g. just firing up chrome will make the AVX offset kick in.

4.7 at 1.245v is still not bad tho.
 
Kirby Wurm said:
If it makes you feel any better mine will only do 4.8GHz at 1.26v with an AVX offset of 2. Without it I have to do 1.3+ and that's after a delid, it really needs a lot of power. There is indeed huge variability.
Click to expand...

What I do to control the power of extreme AVX loads is set a power limit e.g. XTU with no limits at my config would use about 160-180 watts, I have a limit if I remember right of something like 120 sustained, and short term boost to 140 or 150. For all normal workloads this equates to no throttling, games etc. But if I were to start doing rendering on something using AVX instructions it will be contained by the power limit, this manages temps as well.

This way light to moderate AVX loads dont throttle below my max clocks (they would with an offset).
 
You must log in or register to reply here.
Back
Top Bottom