[fateswarm]

Auto-subscribe obviously. Auto-rep obviously. I tend to a UD5H from what I've seen (though I'm still waiting for all boards information and for Devil's Canyon).

 

[fateswarm]

Should we be monitoring voltage with a multimeter if possible in case there is voltage regulation degradation over the months/years? Is voltage regulation degradation a common phenomenon? It came to mind when I noticed people usually blame degradation only to chips.

 

[fateswarm]

Hrm. OK. Then I guess it could be an issue on other brands..

 

[fateswarm]

Quote:

Originally Posted by mandrix

(Not talking about you, fateswarm, I know you're just trying to get to the facts.)

You better not meant I'm on an anti-Gigabyte campaign considering I have one in the mail

 

[fateswarm]

You should add on the guide that max cpu overclocking can benefit from 8x memory. I helped me go from 5.0 to 5.1 on air http://valid.x86.fr/1i9g1v

Though I would be more impressed with very low voltages above 4.6g on high performance settings since it would have a real world benefit.

I currently run 4.6gig, 2400 memory, 40x ring for 24/7 on 1.24v/1.74v (1.20-1.22v had issues).

 

[fateswarm]

I like 8GHz and I'd like to keep the cpu stable for at least a year on 24/7 settings. At 4.7 I get a big jump in voltage requirements. I don't want to go above 1.3v for 24/7 settings.

 

[fateswarm]

Quote:

Originally Posted by Cyro999

[email protected] is pretty normal

Speak for yourself. I want the chip to be stable for a year at least. Also, stop judging others' decisions like they are yours.

 

[fateswarm]

"I'm old here" is not an argument. Find actual arguments. Otherwise you sound like those narcissistic people in the Militaries torturing the newbies.

 

[fateswarm]

Well, let me be blunt. The popular culture of the forum has mixed up what extreme overclockers do with what a 24/7 casual overclock should be for safety in the span of at least ~2 years one would want to use the chip. I have an Intel Engineer on video explicitly saying to go up to 10% above default Vcore (which checks out to be around 1.3v in many cases) and we have reports of degradation in a year (or less) from 1.35v onwards.

 

[fateswarm]

Now you changed the subject. You talked of 1.4v. 1.25v is safe for more than 3 years most likely and 1.35v is probably safe for at least 1-2 years in most cases.

 

[fateswarm]

Quote:

Originally Posted by mandrix

What I said in the DC owners thread bears repeating here:
If you are running your cache voltage on Auto, either check it under load or (probably better option) set it manually. Why?
After testing a recent BIOS for my Z97-UD5H I left the Vring on Auto, ran a quick stress test while manually checking voltages with my DMM and the Vring was 1.397v.
IMO that's a little high.
Fortunately the newer BIOS are finally scaling the uncore from 8x-40x so that's good.

Sin you wanted to know why not having a sensor for vring was disturbing to me? How about longevity? lol.

Definitely. What I wrote in the club:

Quote:

Originally Posted by fateswarm

Ah by the way. The VRing auto-voltage might be also bugged, at least of similar behavior with what VCore had before BIOS updates. I found it above 1.4v in some instances.

It's a bit uncertain how it works because I've also found it lower while on the same BIOS ver. and still on auto and similar multiplier.

It may be affected by other variables.

I meant even on the latest bios.

I'm not certain how it works (or not yet) as it's apparent on the text.

Also I suspect the "quirk" might be that it's treated like OC'ed 4770.

 

[fateswarm]

Thank the gods for multimeter support to be honest. I don't see that voltage on HWInfo.

 

[fateswarm]

ga-z97x-gaming 7 has voltage read points. Ud5h does too. Next to the power button.

Ah you mean on HWInfo? Dunno. But I don't see it here.

 

[fateswarm]

Weird. When I had ring manually on 1.05 (also the default) I couldn't see it on GTL. Now that I've put it on 1.12 I do (though with a weird print of 1.119141 but that's similar on core).

I still don't see it on hwinfo though.

 

[fateswarm]

Is it normal for vcore to be +0.02v than manually set?

On the multimeter. Software reports it as set.

 

[fateswarm]

Quote:

Originally Posted by Cyro999

Everyone went over this a year ago

Sorry I don't know everything Master of the Universe Cyro999.

It appears that on HWInfo, the CPU itself reports it incorrectly as set, and an IT8620E sensor chip reports it close to the multimeter.

It appears that chip kinda connects like a multimeter. They could probably get rid of voltage read points with it on everything.

Hm, the polling rate appears to not be able to go as low.

 

[fateswarm]

Apparently, a few boards report ring too.

 

[fateswarm]

The VRM is downgraded to 4 phases. It's not too bad though since the mosfets are good and it can do i7 K easily. The ethernet card is also intel which ironically is preferred by most.

 

[fateswarm]

It can handle any cpu with its power output. The only thing that it theoretically does worse is voltage stability. But it's likely it won't be noticed or affect anything in most realistic scenarios.

And the capacitance is as high as the Force's if I recall correctly. This may make it similar. Though an electronics expert would be better at answering that.

 

[fateswarm]

Quote:

Originally Posted by hotrod717

Honestly wouldn't think it would make any difference with a 2c pentium.

It probably won't in practice. Not because it's a pentium. On any chip.