5 GHZ i7 6700k?

zLukez

Reputable
Oct 31, 2015
192
0
4,690
So I managed to reach 4.9ghz at approx. 4.2-4.3. I think I can get up to 5 ghz before 1.47. I plan on taking a picture at 1.47 and not benchmarking, so a quick 1 min.
Is this really bad? Also my best preformance is 4.8ghz, 1.39v.
 
Solution
It doesn't mean the oc is bad but there are various limitations that need to be kept in mind. Core voltage limits, temperature limits and stability. It also depends on what you're attempting, do you want a reliable overclock that will allow your cpu to run 3-4yrs+ without killing it or do you want to drop liquid nitrogen cooling on it and push it to a bazillion ghz long enough to set a world record?

It helps to know what scenario you're after. Once you've decided that then it would be helpful to check out some overclocking guides and get a feel for how to go about it. Making small adjustments at a time, testing for temps and stability, making small adjustments, retesting and so on. There are people who push their voltage up higher...
Very unlikely that you'll hit a stable 5GHz, i've seen very few people manage it.
I got a stable 5GHz in one test, but after repeating it, it proved to be very touchy, not worth it for the minimal gain imo, 4.9 was a bit the same, so I just stuck to 4.8GHz, still a fantastic OC.
 
That's pushing it awfully hard. Not to say pro overclockers don't but they know the risks involved and they're prepared to dump another $300 on a new cpu. 1.35v is the preferred upper limit for most for a 24/7 overclock, 1.4v is usually considered the max upper limit of 'safe' and you've already exceeded that.

I'm not sure the point especially if it's 'just to see it on boot' since there's no real accomplishment there. Essentially an overclock that isn't stable, can't pass stability tests, torture tests or that results in a crash as soon as you attempt to do anything beyond viewing the desktop is a 'failed overclock'. At that point you may as well just take a screenshot and alter it in ms paint to say 80ghz because it's just as absurd.

That's why overclocking results submissions for various clubs and things require a screenshot with a hardware monitoring program open, prime95, ibt or other various stress tests open and included on the screenshot to verify. In the end it's up to you, it's your cpu and your investment.
 
It doesn't mean the oc is bad but there are various limitations that need to be kept in mind. Core voltage limits, temperature limits and stability. It also depends on what you're attempting, do you want a reliable overclock that will allow your cpu to run 3-4yrs+ without killing it or do you want to drop liquid nitrogen cooling on it and push it to a bazillion ghz long enough to set a world record?

It helps to know what scenario you're after. Once you've decided that then it would be helpful to check out some overclocking guides and get a feel for how to go about it. Making small adjustments at a time, testing for temps and stability, making small adjustments, retesting and so on. There are people who push their voltage up higher than most of us would, they understand the risk and it's a hobby for them. Usually it's also not their only pc and if they burn up a cpu or motherboard in the process they'll just go buy another.

Sort of like people at the racetrack when they push an expensive engine too hard and blow it up and think oh well, next time. If it's your daily driver to get you to and from work and you don't have the funds to rebuild it should you blow it up, racing it may not be a good idea.

In terms of being a good overclocker or poor overclocker that can be interpreted a few ways. Total ghz to break a record with no attention paid to voltage is one way. Another way is if you're able to achieve a more normal overclock of say 4.7 or 4.8ghz at similar or lower core voltage than other folks with the same clock speeds, that would be a good overclocker. If you crash and have to use a ridiculous amount of voltage to get the cpu 200mhz over stock then it's probably a poor overclocking chip.

Overclocking in steps you'll also see when the cpu begins reaching its limits. If it's only been taking .02 or .05v to reach the next multiplier stable in stress tests and then to reach the next multiplier and still remain stable it requires .1v well that's a large voltage increase just to get another 100mhz out of the cpu. It's indication that the cpu is reaching it's physical limits when small clock increases begin requiring larger and larger voltage increases. It's a point of diminishing returns. That's all assuming that temps are still well within normal ranges, if you're creeping up on 90-100c then you either need better cooling or you need to back off the overclock (and the voltage since the voltage increases heat).

Here's an example of a skylake overclocking guide and it's likely a bit different from haswell guides since baseclock overclocking has come back into the picture.
http://www.overclock.net/t/1570313/skylake-overclocking-guide-with-statistics#

You'll notice some users using voltage of 1.583v vcore which if you ask me is insanely high. Their toys, their choice. They also mention in their overclock documentation that it's part of a project to test degradation so they're well aware of the risk and likely positioned to replace it should it die on them. If you look at the averages of vcore you'll see that very few are over 1.45v. The group of people using 1.45v is fewer than those using 1.4v or below. Everyone has different individual experiences but it makes for a good read through forums like that to get a feel of where other people are at. For example some earlier on were having issues reaching 4.6ghz and staying at 1.4v or under. It gives a relative point of comparison to how well your cpu is doing vs other folks.
 
Solution