Oh boy, this discussion again. Sure, I'll entertain it.
It's common knowledge. Just google it. But here you go:
Common Asrock issue, and it looks like it's holding true on asrock true:
http://www.overclock.net/t/1360404/a...-reading/0_100
Just google "asrock false voltage' i mean jeez, how do you buy a board without googling it?
https://www.google.com/search?q=asro...hrome&ie=UTF-8
http://forums.anandtech.com/showthread.php?t=2308760
http://www.youtube.com/watch?v=oB0dm2-nfpc
Literally, over 30 pages of it here, people saying it:
http://www.overclock.net/t/1360404/asrock-z77-series-vcore-reading
EDIT: !!! THIS SEEMS TO APPLY TO ALL ASROCK Z77 SERIES BOARDS !!!
I just decided to measure the real voltage....
This is really just an issue of common sense, if you know anything about motherboards, all you would have to do is just
look at the Asrock VRM:
no Low RDS on (3 pin with middle leg 'clipped'), single low and single high for each channel, bad drivers, no doubler... i mean you can just see the board is low quality VRM, you don't need anyone to tell you that.
Seriously, go read up on a VRM. Here's a very outdated, old primer for you, that's relevant only because the Extreme4 is that bad:
http://www.hardwaresecrets.com/arti...e-Motherboard-Voltage-Regulator-Circuit/616/4
As you can see, 3 pin mosfets with the middle leg cut out. That means it lacks a modern technology called 'Low RDS ON', which significantly reduces the amount of wasted output on a mosfet when it's in the ON position (ie current is flowing). You will not see non-Low RDS ON on any mosfet made in the last 2 years on any board above $100.
It's really a huge deal these mosfets don't have Low RDS On. This motherboard would literally be called trash 4 years ago back on AM3, that's how bad and outdated Low RDS on is. If you recall, the notorious MSi AM3 boards that were tragically and spectularly catching fire for so many AMD users for Phenom X4/X6, even on stock settings, were using these kinds of Mosfets.
To go more in-depth, these are D-PAK mosfets, which are extremely low quality:
ASRock is using D-PAK MOSFETs under the heatsinks, D-PAKs were phased out years ago on enthusiast VRMs because of their lack of performance compared to newer low RDS(ON) packages such as PowerPAK and LF-PAK when it comes to thermals. They can cause a lot of heat as they aren?t so well suited for really fast switching power supplies such as needed for newer processors.
MSI is the only other company that uses no Low RDS On mosfets, which is why they have a terrible reputation for motherboards, why their boards often catch fire (which now has been fixed by the fact that most of their lower end boards are simply voltage locked or voltage limited), and this is only on their G41 and G43 boards, which cost like $20-50 brand new.
Instead of using a doubler/dual-driver for each 2 phases such as the Z77 Extreme6 does, the Z77 Extreme4 uses 5 Intersil ISL6612 which are just simple drivers(4 for the vCore and 1 for the iGPU). Now ASRock is also doubling its phases, however the issues here is that they are linking two phases to each other, so it is a big 4 phase VRM made to look like a 8 phase VRM with added inductors. There VRM is controlled by an ISL6367, a very high tech analog PWM. We will get into the PWM more with the Z77 Extreme6, as ASRock advertises the PWM as fully digital, while even Intersil doesn?t do this. The Z77 Extreme4?s VRM needs to be a bit better, I mean they could have just used 4 phases of high quality and still had the same cost, but sometimes they have to play the phase wars against GIGABYTE and ASUS. Every phase in my book requires its own driver, thus making this board a 4 phase VRM.
Sin0822, the guy who seems to know the most about VRMs anywhere at the moment
http://forums.overclockers.com.au/showthread.php?t=1062005
Asrock uses Intersil for it's PWM, which is an analogue company. I could go on, and on, but okay, I think I've already established it's a bad board. But what about other boards, or maybe how does this matter in real life? Effectively, this means an Asrock Extreme4 Z77 is going to overheat somewhere around 1.4v+ on Ivy Bridge.
It's very easy to do 1.4v+ on Ivy Bridge, I ran 1.55v 24/7 on Ivy, I know many others ran 1.5v+ with high end air/water, and just about anyone who delidded and/or runs water did 1.4v+. You really need 1.4v+ if you have any intention of running 4.7ghz+ (yes, you could get lucky on your chip, but considering the average voltage for 4.7+).
Now, when it comes to overclocking your CPU, 3 things matter - CPU, MOBO, Heatsink:
1. The Chip. We all know binning matters.. About 50% can only do 4.5-4.6 below 1.45v (3 of my ivies were like that, 4.6@1.4-1.45v). Less than 10% can do 5ghz (my 5'er took 1.55v). But given that Ivy/Haswell increase rather linearly with voltage (not to be confused with heat!), even a bad chip can appreciate a better board+cooler - ie, a 4.5@1.4v would do 4.7@1.55v if you got the cooling.
2. Heatsink. Saying you don't care about your motherboard, or all motherboards are fine, is literally the exact same thing as saying all heatsinks are fine. No, just like you can't do a high overclock without water or at least high end air, you need a high end motherboard to do a high overclock, and a mid-range board to do a moderate overclock. Now a hyper 212+ will do the same thing as a 360mm rad Phobya custom loop w/Helix reservoir if all you do is 4.6ghz@1.3v, but if you want to do a higher overclock, you need a better heatsink. Like, you need a better...
3. Board. A board is literally a power supply to your CPU. Your PSU takes the 120v or 240v from your wall and turns it into 12v (and 5v, and 3.3v, etc). Your motherboard is not only what takes that 12v into ~1v-1.5v, but increases and decreases in nanoseconds based on software/firmware loads (a lot more difficult! that's why high quality PSUs are able to use things like electrolytic capacitors which would literally burst on a modern motherboard).
And just like a power supply, a good motherboard is efficient, it runs cooler, it lasts longer, it provides more stable power (ie a delta based psu like a lower end antec or corsair CX will vary and provide 12.5v, while a capstone will never do more than 12.2v), it has a better transcient (ie it handles changes in load better so you don't crash instantly on load or to idle or sleep because it can't provide the voltage change quick enough or smoothly enough), it performs better in high temps, it has less ripple/noise (ie voltage is more stable so you don't end up using more voltage than necessary), and it has a much longer lifetime.
Finally, a good board will not only supply voltage cleanly, efficiently, and accurately, it will report voltage accurately (no board is accurate, but better boards come at least within .02v of accurate). This is partly hardware, but also firmware, as boards use an algorithm to report software voltage - ie your voltage is never 1.5v, it's 1.4v to 1.6v to 1.4v to 1.6v, this keeps the chip cooler as it uses less voltage as necessary, when necessary, and the fact it'd cost thousands to eliminate this good effect (hence why max LLC is bad for air/water overclocking).
Now the difference between a good board and a bad board is nill when all you do is 4.6ghz@1.3v, but just like a the CPU, a motherboard can overheat. And lower end and mid-range boards will literally overheat before your heatsink do, even with just a mid-range air heatsink. A bottom barrel board like an MSI Z77A-G41 or Pro3 will overheat the same time a Hyper 212+ does, a mid-range board will overheat around the same time a mid-range heatsink like an H50, H60, Kuhler 620, Venomous X, Havik, etc will overheat, and it takes a high end board to keep up with high end heatsinks or water.
For example, my MSI Z77A-G41 on near-stock voltages of 1.25v, would hit over 80C! That's generally when you start to hear coil whine, and yea, it was louder than my 120mm fans on full blast (and a lot more obnoxious, with that high pitch...). 80C is also when you have significant performance degradation - your motherboard efficiency drops below 70% (ie 70% efficiency on 200w...) and you need more voltage for the same level of stability. 90C is when the PCB starts to brown and you see permanent damage over long term use, 100-125C is when things start to blow up in your face on lower quality parts. Meanwhile, my Havik 140 was only at 70C (and it was only 60C on same set-up with UD3H instead, who's VRM ran only 45C at that overclock). The G41 uses a similar MOSFETS to the extreme4. My A770E3 3+1 no low-rds on board hit 90C on just 1.5v on a phenom x4, and that's after I put a sawed off vrm heatsink and fan on it!
On top of a low quality VRM, ALL Asrock Z77 boards completely fudge their vcore numbers! Consistently, every single Asrock Z77 model, for every person who's tested it, is off by over .1v+! The difference between a reported 1.5v and an actual 1.53v really isn't a big deal, but going from 1.45v to 1.6v, that's absolutely insane and extremely dangerous! Only the Extreme4 uses low quality VRM, but all Asrock boards use the same, underreporting algorithm in the firmware to report vcore.
Now every board 'cheats' and this is why motherboard reviews and benchmarks are so damn misleading and bad. Every motherboard company fudges things a bit, all of them. Gigabyte dynamically loosens and tightens one of the tertiary RAM timings, so it magically scores higher in RAM overclocks and benches. Asus dynamically overclocks the BCLK, so it appears to run faster than other boards at stock when in reality it's not 'stock' at all. MSI, Biostar, they all do this. Asrock's little cheat, is to fudge the vcore, which hurts overclockers... but helps sales.
You might think your Asrock Extreme4 is fine at 4.7ghz@1.35v, but truth of the matter is that the board will die in ~5 years instead of 10+ years that a similarly priced Gigabyte, Asus, Biostar, or even MSi board will do. Furthermore, your VRM is getting hotter, meaning you run way less efficiently, meaning you are spending an extra $10/year in energy wasted (hotter temps = higher power consumption, 10C is about 10-20w for a 150-200w VRM and a 150-200w CPU). Your chip is also getting hotter, so you need better cooling than you would need otherwise.
Literally, an Asrock p67/z68 is a higher quality board than their z77 boards. Hence, the Extreme4 was widely purchased. Furthermore, there were a
ton of extremely misleading reviews on the Asrock Extreme4 (and asrock series in general), inciting tons of people to buy the Asrock Z77 boards when in reality they were the worst choice. In fact, here's a choice page from a very bad offender, written by a terrible reviewer:
http://www.anandtech.com/show/5793/...ith-ivy-bridge-asrock-asus-gigabyte-and-msi/3
This reviewer talks about how great the Asrock board was, when in reality he wasn't even using a multimeter,
he used software to measure vcore which we all know is a cardinal sin. Furthermore, he is using CPU-Z (via OCCT), which back then was buggy in it's temp readings and specifically altered these results. He also talks about vdroop and ripple, which is absolutely useless and absurd - you will never see vdroop or ripple using software voltage reports, by their very nature,
software reports an average, meaning the droops and peaks and valleys, will be left out!
If all you do is mild overclocks, don't waste your money on an Extreme4 when a $20 MSI Z77A-G41 would do the same job. Conversely, an Extreme4 is incapable of high end overclocks, and will suffer more on moderate overclocks, than the comparably priced Biostar, Gigabyte, and Asus boards. Mid-range Gigabyte boards on Z77, particularly D3H and up, offer high end quality for mid-range price, and the Asus boards use a decent VRM for their price. I'm sure for most of you the Extreme4 'works' just fine, but the issue is that something significantly cheaper would work the exact same, and something the same price would work 2-3x better and last 2-3x longer.
Finally, and so least importantly, is the motherboard build quality, connections, sound chip. Asrock uses same sound and connections as everyone else at the price level, but the Asrock's boards have absolutely atrocious build quality. It's not really a huge deal, but hold an Extreme4, and then hold a UD3H, and the difference is night and day. Also, the Extreme series boards have a reputation for being flimsy and weak, and the extreme4 is also slightly smaller than comparable boards. You can just see the board is 'empty' - it has the same sound chip, but is lower quality due to lower quality components powering the sound. There's little things like this all over the board, for the usb, sata, etc, that are all slightly lower quality on the Extreme4 (it may be slightly slower, likely completely unnoticeable in anything but benches if that, won't last as long, not as durable electronically). They are all quite minor, but basically Asrock Extreme series boards have terrible build quality and are flimsy, and things like sata connectors, usb, etc, really aren't very important on a board nor very different from one board to the next. I've handled and built a few builds with Extreme4s, and they definitely are flimsy pieces of crap in comparison (they feel like $60 biostar or msi boards).
You constantly hear people saying 'don't cheap out on the PSU' or 'you need a good heatsink for a good overclock', but it blows my mind that people overlook the motherboard when it's more influential than a PSU and overheats just like CPU. Don't tell me 'modern CPUs don't use that much power', because an i7-3770K can easily pull 200w+ at 5ghz@1.45v (mine did, killawatt, software, check sin's ib oc guide), which means your motherboard is pulling well over 250w due to (in)efficiency and VTT/IMC/iGPU/etc.
To blow off the board is MUCH sillier than blowing off a PSU, a $20 cx430 can still be used for high end overclocks just fine, and tell me again how little energy Ivy/Haswell uses yet still require 200w+ heatsinks. All that heat, all that wattage, emitted by your CPU, runs through your motherboard first!