Question about OCing and Voltages

[DarkRogue]

Hello, everyone,

I'm trying to decide what motherboard to get for an upcoming IVB build, and was looking at either the Asus Z77 V Deluxe, or Gigabyte's UD5H. I'll be pairing it with a 3770K and looking to OC into the 4.5GHz~ range, which seems reasonable so far.

Anyway, so I've been looking up information about these boards with relation to overclocking, and I'm finding myself puzzled at the voltages being reported by CPU-Z.

First off, the ASUS board sports more power stages or something, which sounds like it'd be helpful to overclocking.

However, thus far, every article I've found seem to have the ASUS board reporting much higher voltage via CPU-Z than Gigabyte boards for very similar overclocks.

Is this a problem with CPU-Z misreporting, or do the ASUS boards really require more voltage to be pushed through? Is it a result of having more power phases? I thought OC'ing would be more dependent on the CPU itself, where it'd have similar settings regardless of motherboard, but there are differences of up to 0.3v between the two boards.

I'm confused.

 

[DarkRogue]

Anyone? :/

It just occurred to me that I didn't check for temperature data, to see if it was simply a mis-reporting of the voltage somewhere, or maybe CPU-Z showing differences in how the mobos handle vdroop.

I'm trying to double back and check now, but many of these OC reviews don't have temperature data attached. Ugh.

 

[T_Yamamoto]

All boards will have their ups and downs. Since ivy bridge isn't officially out, we won't know which board is the best yet. Wait a little while but I would highly recommend an ASRock board

 

[dma0991]

It should be the same. Probably you're looking at reviews where they happen to be using a slightly better binned chip than the rest. My Core i5 3570K is somewhat similar to the one that was tested by Anand where I'll be doing around 4.5GHz at 1.28V (1.295V manual set, 1.280V after vdroop).

Some Core i7 3770K's I've seen so far are able to do that are capable of doing nearer to the 1.2V mark and stable. Under 1.3V for 4.5GHz is easily achieved for any IB and you're still within its safe range. Go any higher than 1.3V and 4.5GHz and you'll see the voltage requirement will spike along with the temp increase.

 

[Charlie98]

My Gigabyte board erroneously reports VTT as 'core voltage.' It fooled me at first until I verified it with the Gigabyte EasyTune6 voltage monitor utility.

 

[MPiland]

Hello, everyone,

I'm trying to decide what motherboard to get for an upcoming IVB build, and was looking at either the Asus Z77 V Deluxe, or Gigabyte's UD5H. I'll be pairing it with a 3770K and looking to OC into the 4.5GHz~ range, which seems reasonable so far.

Anyway, so I've been looking up information about these boards with relation to overclocking, and I'm finding myself puzzled at the voltages being reported by CPU-Z.

First off, the ASUS board sports more power stages or something, which sounds like it'd be helpful to overclocking.

However, thus far, every article I've found seem to have the ASUS board reporting much higher voltage via CPU-Z than Gigabyte boards for very similar overclocks.

Is this a problem with CPU-Z misreporting, or do the ASUS boards really require more voltage to be pushed through? Is it a result of having more power phases? I thought OC'ing would be more dependent on the CPU itself, where it'd have similar settings regardless of motherboard, but there are differences of up to 0.3v between the two boards.

I'm confused.

From a few of the reviews I read, the ASUS board seems to come out on top with the AsRock boards not far behind. I went with the P8Z77-V Pro now just waiting on my chip. I'll let you know how it works out.

 

[Concillian]

1) For a moderate OC at 4.5 GHz, you do not need to be concerned with special overclocking features like power phases. They do provide a small amount of margin, but at a moderate OC, these won't really come into play. You don't need an expensive motherboard just for OCing. Buy the expensive motherboard for the additional ports better layouts, and other non-OC related features.

2) Reviews consistently show the ASUS boards higher on load power draw as well. (example: http://www.guru3d.com/article/asus-maximus-v-gene-motherboard-review/8 the two ASUS boards are 20+watts higher than the MSI and Gigabyte board, and I've seen other reviews showing similar results).

This could be that the extra voltage is real, or it could be power to provide for all the misc. stuff that ASUS adds to boards to differentiate themselves from other vendors too.


Regardless, for a 4.5GHz OC, you probably don't need to worry about anything you're worrying about. The mobo isn't going to make a scrap of difference on that OC, it's all up to your luck in the silicon lottery and your cooling, but mostly your lottery ticket.

 

[DarkRogue]

Whew, I stayed up far later than I should've, researching this crap, lol.

Yamamoto, yes, all mobos do have their differences, which is why I was trying to ask for clarification. IVB isn't officially out for the masses yet, but the NDA has lifted, so there are lots of reviews available, as well as people "sourcing" CPU's from somewhere. Otherwise, with respect to my question about vcore, I believe it should hold true with the tests done with SB as well.

dma0991, I thought about those differences as well, but throughout my searching, the difference was consistent. It's very unlikely that everyone who tested a Gigabyte mobo somehow got a higher binned CPU that ran with less voltage. Furthermore, LegitReviews used the same CPU (I believe) on both mobos, and it still saw a similar difference.

MPiland - please let us know how it turns out for you!

Concillian, I had noticed as well that the ASUS boards consumed more power with all else being equal as well. I'm not sure what exactly is causing that, but it could be as you say; more power used for added features.

Charlie98, I think you're on the right track. It's possible that either CPU-Z or Gigabyte is not understanding what the other needs.

From what I found last night, while I couldn't find many reviews that posted temperatures of their overclocks, a few of them did say what voltage they had set, which I then compared with the CPU-Z screenshots.

Most overclocks on the 3770K on these two motherboards were within the rough 1.3v~ range. However, the Gigabyte CPU-Z screenshots consistently showed voltages that were under 1.1v where the Asus motherboards had screenshots showing voltage that were much closer, around 1.3~1.4v.

I don't know if there was anything to do with load line calibration on either motherboard, or if they were run during idle/load and maybe vdroop occurred or not, though.

Suffice to say, I'm still quite confused, but at least I have some information and a possible direction to pursue.

 

[Concillian]

Most overclocks on the 3770K on these two motherboards were within the rough 1.3v~ range. However, the Gigabyte CPU-Z screenshots consistently showed voltages that were under 1.1v where the Asus motherboards had screenshots showing voltage that were much closer, around 1.3~1.4v.

Those are not on the exact same CPU, are they?

That much voltage difference can't be due to the motherboard. That's got to be the difference between great silicon and mediocre silicon. There is always significant variation between the good and bad chips.

As an example Guru3d has so far tested 3 boards at 5GHz with the same i7-3770k, and the voltage is only varying between 1.416 to 1.48 on the CPUz screenshots. changes from 1.1 to 1.3 are not likely to be motherboard issues.

 

[DarkRogue]

Unless review sites had access to more than one CPU (in which case, I would be heavily mistaken) I'm pretty sure reviewers used the same CPU for multiple motherboards.

Otherwise, across sites, yes the CPU's are all different.
However, I wasn't looking for absolute values, so much as I was looking for a trend, and the trend I found was that CPU-Z screenshots posted with Gigabyte mobos were always reporting much lower voltages.

I'll post a few examples from my research; please correct me if my methodology is incorrect here.

First, LegitReviews is my biggest example, because they reviewed both mobos that I was interested in at the same time, so I assume they used the same CPU.

Source: http://www.legitreviews.com/article/1882/16/
Asus P8Z77-V Deluxe & 3770K @ 4.8GHz
They stated that they set the vcore to 1.35v
CPU-Z:

Gigabyte Z77X-UD5H & 3770K @ 4.8GHz
They didn't state what vcore they used this time, but noted that they couldn't OC any more due to high temps and "high vcore"
CPU-Z:

Source: http://www.overclock3d.net/reviews/cpu_mainboard/gigabyte_z77x-ud5h_review/6
Gigabyte Z77X-UD5H & 3770K @ 4.8GHz
They stated they used a 1.32v vcore
CPU-Z:

Source: http://www.eteknix.com/reviews/motherboards/gigabyte-z77x-ud5h-z77-motherboard-ivy-bridge-review/3/
Gigabyte Z77X-UD5H & 3770K @ 4.9GHz
They stated that they needed 1.425v on their vcore
CPU-Z:

As far as the ASUS goes..

Source: http://www.kitguru.net/components/cpu/zardon/asus-p8z77-v-deluxe-review/5/
ASUS P8Z77-V Deluxe & 3770K @ 4.6GHz
They stated that it required 1.24v and "CPU Power Duty Control" (LLC?) to 'extreme'
CPU-Z:

They further pushed it to 4.8GHz with 1.33v
CPU-Z:

Source: http://www.tweaktown.com/reviews/4674/asus_p8z77_v_deluxe_intel_z77_motherboard_review/index4.html
ASUS P8Z77-V Deluxe & 3770K @ 4.7GHz
No vcore stated, unfortunately.
CPU-Z:

Not an exhaustive list by any means, and I'm sure I've forgotten a couple that I went through, but it seems to me that the CPU-Z screenshots appear to show voltages that are closer to what users used on the ASUS boards, and much further from the Gigabyte boards.

 

[Charlie98]

Here's what I'm talking about...

 

[Concillian]

Yeah, those 1.0x voltages are clearly not real, you aren't going to see one board hit 4.8 GHz at 1.35 and another board hit it at 1.07. Motherboards can make a difference, but not THAT much of a difference.

 

[Maximilian]

Yeah gigabyte boards show the VTT voltage as the vcore in cpu-z and hwmonitor for some reason. No idea why this is.

 

[DarkRogue]

Thanks guys, I guess it's an issue with Gigabyte boards misreporting after all, then.

Curious no one else thought to mention that anywhere, as if it was supposed to be common knowledge.

Does Gigabyte only do this on these z77 boards, or did they always do it regardless of chipset? I don't even remember what I set my vcore to anymore.

 

[Charlie98]

Well, that image is with my Gigabyte Z68 board, but I suspect it will be the same with Z77, etc.

Also, you mentioned not knowing what your vCore is... you need to find out.

I do NOT recommend using any of Gigabyte's OC utilities (Smart6 or EasyTune6,) from what I've seen, they pump up the BCLK along with the multiplier... and SB doesn't, again, in my experience, like BCLK higher than 101.0. But, as you can see, I still use the EasyTune6 utility to monitor my voltages, it works very well.

I had to go in and reset my CMOS, reenable AHCI, and start over once I realized what those utilities were doing...