Question Is this a safe 5.2ghz on a 10700k?

[Rodrigodrt]

Vcore is 1.38v @ 5.2Ghz
with LLC2 on a MSI I get a a bump to 1.384v when stress testing on aida/cpu-z/heavy games, managed 1 hour no crash, but im uncertain if its REALLY stable, so i may need to bup .05 or .1 if i really have a wiggle room for 1.4v maximum, i wouldnt go past that.

on gaming the peak temps are 69 - 75 ish on most games, but 2 games are temps hungry, raft and no mans sky can reach low 80~83c peaks, averaging 76 - 83C on peak time.

Is this safe? I don't think i have a single digit room for lower vcore, or higher LLC, this seems to be the limit for 5.2, and I really like that number, but don't wanna play on dangerous side. I hear a lot 1.35v is max and 1.4v is actually the max safe, hell even 1.45v if kept cool. What do you guys think?

 

[Deleted member 362816]

Overclocking is never "safe" that being said your Vcore isn't to bad and your temps are fine.

 

[Rodrigodrt]

Overclocking is never "safe" that being said your Vcore isn't to bad and your temps are fine.

well yes, but often OC'ers adopt this or that 'safe' temp and voltage limits so that you get to run your chip for a couple of years, enough so that if any degradation happens you're already bound to upgrade anyways. I'm trying to find that just perfect spot. but if my temps and oc looks fine, im more reliefed since i come from Haswell, and those the safe is generaly up to 1.35v absolute max, yet 10 gen have supporters ranging from 1.35v to 1.45v like i said, so..

 

[hotaru.hino]

"Safe" is whatever is within the specifications of the part, which you can find at https://www.intel.com/content/www/u...r-16m-cache-up-to-5-10-ghz/docs.html?s=Newest. Looking through the documents, it lists the maximum voltage as 1.52V as a base, up to 1.7V with an offset.

There's also Intel's official overclocking guide: https://www.intel.com/content/www/us/en/gaming/resources/how-to-overclock.html. While they warn to never go above 1.4V, it's more out of a concern of how much cooling you can do than any risk of outright damaging the part. You can see later down in the article they have a graph showing a processor being pushed to 1.5V.

If you want a comfortable spot, just don't go above 1.4V and do whatever helps you sleep at night.

 

[Roland Of Gilead]

Vcore is 1.38v @ 5.2Ghz
with LLC2 on a MSI I get a a bump to 1.384v when stress testing on aida/cpu-z/heavy games, managed 1 hour no crash, but im uncertain if its REALLY stable, so i may need to bup .05 or .1 if i really have a wiggle room for 1.4v maximum, i wouldnt go past that.

on gaming the peak temps are 69 - 75 ish on most games, but 2 games are temps hungry, raft and no mans sky can reach low 80~83c peaks, averaging 76 - 83C on peak time.

Is this safe? I don't think i have a single digit room for lower vcore, or higher LLC, this seems to be the limit for 5.2, and I really like that number, but don't wanna play on dangerous side. I hear a lot 1.35v is max and 1.4v is actually the max safe, hell even 1.45v if kept cool. What do you guys think?

The general consensus is to have your max temps (stress load - not gaming!) at or below 80c for 24/7 OC. If you are hitting those temps in gaming, chances are if you stress the CPU with Prime or something, it will crash or hit really high temps. It won't be stable. you should consider stronger cooling.

You need to dial it back a little. Some will argue you are leaving performance on the table by using Prime or similar as a measure. However, I'd much rather my OC be stable all day long at 5ghz, than potentially unstable with 5.2ghz. The 200mhz will yield practically nothing in terms of FPS given your high specs. You would only be chasing numbers for synthetic benchmarks with 5.2 over 5.0 or even 4.9.

Just reduce the multi two notches, and keep below 1.375 vcore. Your LLC at level 2 has reduced vdroop nicely, so keep that.

edit: You might get lower than 1.375 at 5ghz, if you fine tune. I was just using that as it's typically close to the max VID for those chips.

 

[Phaaze88]

Just take it from one of Intel's own engineers: https://www.tomshardware.com/features/inside-intels-secret-overclocking-lab/7

You likely need to dial it back a bit.

 

[storm-chaser]

Just take it from one of Intel's own engineers: https://www.tomshardware.com/features/inside-intels-secret-overclocking-lab/7

You likely need to dial it back a bit.

Okay but this is a quote taken directly from that write up:

Speaking as enthusiasts, the engineers told us they feel perfectly fine running thier Coffee Lake chips at home at 1.4V with conventional cooling, which is higher than the 1.35V we typically recommend as the 'safe' ceiling in our reviews. For Skylake-X, the team says they run their personal machines anywhere from 1.4V to 1.425V if they can keep it cool enough, with the latter portion of the statement being strongly emphasized.

At home, the lab engineers consider a load temperature above 80C to be a red alert, meaning that's the no-fly zone, but temps that remain steady in the mid-70’s are considered safe. The team also strongly recommends using adaptive voltage targets for overclocking and leaving C-States enabled. Not to mention using AVX offsets to keep temperatures in check during AVX-heavy workloads.

 

[Eximo]

Well, Intel employees do have easy access to replacement chips, and they likely don't keep them long.

I ran my i7-7700k at 1.416 volts for several years, it had spikes into the low 80s as well, but did run 5Ghz all core. (I did re-lid it with normal thermal paste to get the core temperatures more even, one was almost 10C hotter, only 3C hotter afterwards) I've removed the overclock, and it is still operating as my nephew's gaming box to this day.

14nm node is quite mature though, so probably safe to stick with the 1.4 volt limit.

I'll have to check the voltage my 10900F reaches when it hits 5.2 and 5.1Ghz, can't quite recall.

 

[storm-chaser]

I run my 9600KF at 1.4v and have no problems. As good as the day I bought it!

CPU degradation is more of a myth than anything else. So long as you have proper cooling, most normal OC enthusiasts will never degrade their CPU by running slightly more voltage.

 

[Deleted member 362816]

I run my 9600KF at 1.4v and have no problems. As good as the day I bought it!

CPU degradation is more of a myth than anything else. So long as you have proper cooling, most normal OC enthusiasts will never degrade their CPU by running slightly more voltage.

Most OC enthusiasts wouldn't know as they usually jump ship in 6months-1 year for a new CPU. That being said the CPU will be irrelevant before 1.4V with good cooling kills it.

 

[storm-chaser]

Most OC enthusiasts wouldn't know as they usually jump ship in 6months-1 year for a new CPU. That being said the CPU will be irrelevant before 1.4V with good cooling kills it.

I have a number of overclocked CPUs that have lasted many years in OC configuration with substantially more voltage than stock.
Phenom II, Core 2 duo, Core 2 quad, Xeons, haswell, ivy bridge, coffee lake, etc

In fact I am a more aggressive overclocker than most and not once have I ever had a CPU degrade, Guess that's why I don't buy into it as much as some do.

 

[Phaaze88]

Okay but this is a quote taken directly from that write up:

*Speaking as enthusiasts, the engineers told us they feel perfectly fine running thier Coffee Lake chips at home at 1.4V with conventional cooling, which is higher than the 1.35V we typically recommend as the 'safe' ceiling in our reviews. For Skylake-X, the team says they run their personal machines anywhere from 1.4V to 1.425V if they can keep it cool enough, with the latter portion of the statement being strongly emphasized.

At home, the lab engineers consider a load temperature above 80C to be a red alert, meaning that's the no-fly zone, but temps that remain steady in the mid-70’s are considered safe. The team also strongly recommends using adaptive voltage targets for overclocking and leaving C-States enabled. Not to mention using AVX offsets to keep temperatures in check during AVX-heavy workloads.*

From that combined with the OP's reported thermals - which they weren't very clear on - just gaming stuff was reported, and that doesn't really count, IMO.
Common causes for temperature spikes in games is caused by the gpu(a 3080 in there), Load Line Calibration level, and momentary utilizations of AVX instructions.

The engineers recommend using adaptive voltage. OP is using fixed. Probably doesn't make a big difference whichever is used, but the latter is quicker and easier to use.

If the OP ran:
-Aida FPU only
-OCCT's small data set, steady load, SSE instructions
-Prime 95 small FFT, all AVX options off
They'd be 1 for 2 from that write up: still have room to go higher on the voltage, but were probably pushing over 80C in those stress tests before doing so. That was the reason for my first post.


OP hasn't been here in over a week... they might've abandoned thread.