F5F Stay Refreshed Power Users Overclocking This Intel 8700k OC 1.42v chip runs at 5ghz and is considered less risky than commonly believed.

This Intel 8700k OC 1.42v chip runs at 5ghz and is considered less risky than commonly believed.

This Intel 8700k OC 1.42v chip runs at 5ghz and is considered less risky than commonly believed.

_
_MrJunior
Junior Member
17
10-12-2017, 05:45 AM
#1
That's a statement, not a question so be mindful of that folks. I more so wanted to open a dialogue about voltages you push through your CPU's to hit certain frequencies. For some reason many people think that a majority of OC'ers can hit the golden 5.0ghz at 1.35vs or less. It is my opinion that anything less that 1.35v would be a lottery item and the exception rather than the rule as I see so much on this forum as well as others. I think too many people look at the voltages alone and not enough of them actually looks at the temps under load which is what matters. I think a majority of chips out there for the 8700k for example, will need 1.38-1.42 from what I can tell, in order to run a 5ghz OC. I can push my own chip at 1.42v for 5ghz and run 35 degrees Celsius at idle with 75 degrees under top load during Prime95. I am sure most people would tell me to step my frequency down to 4.8-4.9 to back the voltages down but I honestly wont worry too much about it. I got the absolute crap side of the lottery as I cant get a stable 5ghz at less than 1.4v and it becomes most stable at 1.42v.
What is everyone's thoughts on this? I know there will be people who say that you will degrade the chip over time running at that high of a voltage and perhaps that holds true. But if it takes roughly 3-5 years to become crippled, wouldn't it be time to upgrade the processor anyway?
I welcome all discussion and thoughts on the matter. You are all awesome people.
_
_MrJunior
10-12-2017, 05:45 AM #1

That's a statement, not a question so be mindful of that folks. I more so wanted to open a dialogue about voltages you push through your CPU's to hit certain frequencies. For some reason many people think that a majority of OC'ers can hit the golden 5.0ghz at 1.35vs or less. It is my opinion that anything less that 1.35v would be a lottery item and the exception rather than the rule as I see so much on this forum as well as others. I think too many people look at the voltages alone and not enough of them actually looks at the temps under load which is what matters. I think a majority of chips out there for the 8700k for example, will need 1.38-1.42 from what I can tell, in order to run a 5ghz OC. I can push my own chip at 1.42v for 5ghz and run 35 degrees Celsius at idle with 75 degrees under top load during Prime95. I am sure most people would tell me to step my frequency down to 4.8-4.9 to back the voltages down but I honestly wont worry too much about it. I got the absolute crap side of the lottery as I cant get a stable 5ghz at less than 1.4v and it becomes most stable at 1.42v.
What is everyone's thoughts on this? I know there will be people who say that you will degrade the chip over time running at that high of a voltage and perhaps that holds true. But if it takes roughly 3-5 years to become crippled, wouldn't it be time to upgrade the processor anyway?
I welcome all discussion and thoughts on the matter. You are all awesome people.

C
Chiller9592
Senior Member
670
10-15-2017, 09:53 AM
#2
It varies depending on the individual's needs. I still notice many users with first or second generation i7s who boost the CPU beyond its recommended voltage, which only slightly affects gaming performance. It seems illogical to me if lowering the lifespan to about five years or less were possible, especially since a properly maintained CPU should last much longer.
C
Chiller9592
10-15-2017, 09:53 AM #2

It varies depending on the individual's needs. I still notice many users with first or second generation i7s who boost the CPU beyond its recommended voltage, which only slightly affects gaming performance. It seems illogical to me if lowering the lifespan to about five years or less were possible, especially since a properly maintained CPU should last much longer.

U
uHP
Member
53
10-15-2017, 06:10 PM
#3
Very good observation. I hadn't considered the people who keep their older generation processors with their original chip designs. Although I can't represent the technical specifications for those earlier models, the 8th Generation 8700K Intel chips specify a maximum voltage of 1.52 volts. Clearly, you should never exceed this unless you have robust cooling, and even then... to my view, 1.4 volts isn't a major concern and shouldn't cause as much worry as some believe. My main point is that the idea of keeping voltage at or below 1.35 seems outdated now and could deter people aiming for 5GHz performance out of fear of damaging their devices.
U
uHP
10-15-2017, 06:10 PM #3

Very good observation. I hadn't considered the people who keep their older generation processors with their original chip designs. Although I can't represent the technical specifications for those earlier models, the 8th Generation 8700K Intel chips specify a maximum voltage of 1.52 volts. Clearly, you should never exceed this unless you have robust cooling, and even then... to my view, 1.4 volts isn't a major concern and shouldn't cause as much worry as some believe. My main point is that the idea of keeping voltage at or below 1.35 seems outdated now and could deter people aiming for 5GHz performance out of fear of damaging their devices.

B
baris070x
Junior Member
49
10-17-2017, 05:05 AM
#4
I operate at 1.4V and rely on the Intel Temperature Guide by CompuTronix. https://forums. This is the most thorough guide I've encountered for Intel CPUs, with significant effort put into it and kept current. It's a great resource when you have the time.
B
baris070x
10-17-2017, 05:05 AM #4

I operate at 1.4V and rely on the Intel Temperature Guide by CompuTronix. https://forums. This is the most thorough guide I've encountered for Intel CPUs, with significant effort put into it and kept current. It's a great resource when you have the time.

M
Mod_masta
Member
191
10-19-2017, 09:14 AM
#5
Is it 75ºC using Prime95 version 29.8, a small FFT test, with AVX2 active, lasting 15 minutes? If yes, that's an impressive performance.
M
Mod_masta
10-19-2017, 09:14 AM #5

Is it 75ºC using Prime95 version 29.8, a small FFT test, with AVX2 active, lasting 15 minutes? If yes, that's an impressive performance.

H
Hermie12
Member
57
10-19-2017, 03:11 PM
#6
The instructions for Avx on small FFTs are completely impractical and unnecessary. I’d recommend running a basic FFT without Avx—it’s sufficient for a quick check.
H
Hermie12
10-19-2017, 03:11 PM #6

The instructions for Avx on small FFTs are completely impractical and unnecessary. I’d recommend running a basic FFT without Avx—it’s sufficient for a quick check.

S
218
10-19-2017, 06:52 PM
#7
Respectfully, based on the information given, I don't see how that's possible.
Many users tend to fling numbers around like Gorilla poo in a cage, without providing the background details such as ambient temperature to support their claims. Without the details, load and idle Core temperatures have no point of reference. Please fill in the blanks and be more specific:
Ambient temperature?
Cooling?
Delidded?
Prime95 version?
Torture test?
AVX?
AVX Offset?
Power consumption (watts)?
• Silicon Lottery
:
Historical Binning Statistics
-
https://siliconlottery.com/pages/statistics
S
shacklebolt323
10-19-2017, 06:52 PM #7

Respectfully, based on the information given, I don't see how that's possible.
Many users tend to fling numbers around like Gorilla poo in a cage, without providing the background details such as ambient temperature to support their claims. Without the details, load and idle Core temperatures have no point of reference. Please fill in the blanks and be more specific:
Ambient temperature?
Cooling?
Delidded?
Prime95 version?
Torture test?
AVX?
AVX Offset?
Power consumption (watts)?
• Silicon Lottery
:
Historical Binning Statistics
-
https://siliconlottery.com/pages/statistics

L
Lucilu123
Junior Member
47
10-19-2017, 08:01 PM
#8
So, to answer your question
@alceryes
, yes. That was 30 mins of Prime95 with the above specifications with 1.42v for 5ghz. However, after making this post this morning I wondered if my Ram was causing some destabilization. My ram I used under the above specs was a pair of Ballistix 2400mhz 16GB (2x8GB). My system would crash just trying to boot up until I got to about 1.38v and then wouldn't be 100% stable running stress tests until 1.42v. After work I purchased a set of Corsair Vengeance Pro RGB @ 3000mhz 16GB (2 x 8GB). After installing the new ram and running an XMP profile (which bumps that Corsair from 2400mhz to 3000mhz, I could boot as low as 1.30v (though not remotely stable). I haven't done extensive testing but I know for sure I can run all stress tests at 1.38v now and be stable after 30 mins. Obviously my temps have lowered since pulling back from 1.42v. Was pretty cool! Idk why I didn't think of the Ram as being something that could hold back an OC but I did.
L
Lucilu123
10-19-2017, 08:01 PM #8

So, to answer your question
@alceryes
, yes. That was 30 mins of Prime95 with the above specifications with 1.42v for 5ghz. However, after making this post this morning I wondered if my Ram was causing some destabilization. My ram I used under the above specs was a pair of Ballistix 2400mhz 16GB (2x8GB). My system would crash just trying to boot up until I got to about 1.38v and then wouldn't be 100% stable running stress tests until 1.42v. After work I purchased a set of Corsair Vengeance Pro RGB @ 3000mhz 16GB (2 x 8GB). After installing the new ram and running an XMP profile (which bumps that Corsair from 2400mhz to 3000mhz, I could boot as low as 1.30v (though not remotely stable). I haven't done extensive testing but I know for sure I can run all stress tests at 1.38v now and be stable after 30 mins. Obviously my temps have lowered since pulling back from 1.42v. Was pretty cool! Idk why I didn't think of the Ram as being something that could hold back an OC but I did.

R
Rinusvandijk
Member
141
10-19-2017, 08:25 PM
#9
I have a portable AC unit in my gaming room that cools down to between 72-74 degrees, which aligns closely with the 72 used by Intel during their tests.
I performed the blend test, expecting results similar to the above, and confirmed it works with AVX.
Although I consider this test unrealistic for normal usage, I’m willing to push it to 5GHz.
If I were aiming for 5.1GHz or higher, I’d be surprised, but I don’t think anything below that would require such performance.
My cooling system is the Corsair 240mm H100I Pro.
R
Rinusvandijk
10-19-2017, 08:25 PM #9

I have a portable AC unit in my gaming room that cools down to between 72-74 degrees, which aligns closely with the 72 used by Intel during their tests.
I performed the blend test, expecting results similar to the above, and confirmed it works with AVX.
Although I consider this test unrealistic for normal usage, I’m willing to push it to 5GHz.
If I were aiming for 5.1GHz or higher, I’d be surprised, but I don’t think anything below that would require such performance.
My cooling system is the Corsair 240mm H100I Pro.