UD3H vs. UD5H Overclocking Potential

ericjohn004

Honorable
Oct 26, 2012
651
0
11,010
A couple weeks ago I made a thread asking whether a Gigabyte UD5H would be better for overclocking than a UD3H. I asked this because my UD3H broke and I had to replace my MoBo and I was thinking of switching to the more expensive UD5H. I do have a question I would like someone to answer for me at the end of my little review. But let me first get to my results that might suprize some of you. Overclocking on the UD5H gave me about a .2Ghz boost in how much further I could overclock because of how much lower I was able to set the voltages. And thus how much lower my temps inevitably were.

This chart tells you what voltage I had to have to reach a certain overclock on the UD3H vs. the UD5H.

UD3H 4.5Ghz 1.300 Vcore UD5H 4.5Ghz 1.255 Vcore
UD3H 4.6Ghz 1.340 Vcore UD5H 4.6Ghz 1.285 Vcore
UD3H 4.7Ghz 1.390 Vcore UD5H 4.7Ghz 1.330 Vcore
UD3H 4.8Ghz 1.420 Vcore UD5H 4.8Ghz 1.370 Vcore

For every .1Ghz I notice my temps move up or down 1-3C. So since the UD5H generally lets me set a whole .05 lower Vcore sometimes a little more, I can then get from 5-15c lower in temps. Allowing me to bump my overclock up to 4.7Ghz from 4.5Ghz while having the same temperature.

I just wanted to let some of you know the difference it could make by going with a more expensive board. The UD3H costs about 140$ whereas the UD5H costs about 180$. Some people talk about hitting the silicone lottery with a certain CPU. But maybe it's slightly less about the chip and a little more about the quality MoBo they are using. As before I'd say I had an average 3570k, now it seems I have quite a good one, although the MoBo is all that's changed.

My question is, have any of you had different MoBo's with the same CPU and had such a big difference in the voltage it actually took to achieve an overclock? I thought this was quite an interesting question. Hopefully some of you think so too and I might get a few replys.

I made an edit changing some of my voltages I have listed for the UD5H. After playing with it a little I've achieved overclocks with notibly lower voltages. I played with the UD3H just as much and the numbers listed are the best I could achieve.

Why am I getting these better overclocks just by spending 40 more dollars on a MoBo???
 

ericjohn004

Honorable
Oct 26, 2012
651
0
11,010


I'm an avid over clocker and I have overclocked and owned 3 Ivy Bridge based QUAD core processors myself, so I know them well. They are usually VERY equal in over clocking despite what some may say, as long as you use the exact same settings. They're may be outliers that require way more voltage but this is extremely rare. The reason I know this is because my friend has built and sold around 400 PC's. Each of them came with a 2500k, 2700k, 3570k, or 3770k. And every one of those 400 CPU's were overclocked to 4.5Ghz while requiring no more than 0.05v differences(the Ivy Bridges not Sandy). Of course, they all used the same motherboard. Which is very important for overclocking and different mobo's will require different voltages. So if everything else is the SAME, all Ivy's should overclock within 5% of eachother 99.5% of the time.

Now, let me just say that there's no way your at 4.5Ghz with only 1.140v and your completely stable. You may be able to boot into Windows and run and hour of Prime95 and some benchmarks but that doesn't mean your stable. The least I would say a 4.5Ghz overclock would require is 1.200v or more be COMPLETELY stable. It takes me 1.265v to be completely stable while most chips are between 1.250v-1.300v.

Are you running Prime95 or Intel Burn In Test with this overclock? At least 3 hours or Prime95 and 150 passes of IBT are REQUIRED to make sure your stable. I've done plenty of testing and know this as fact. Some say you don't have to run a stress test or hours of Prime95 to see if your stable. They're dead wrong. As one day they'll be gaming and will not have saved and they'll get a BSOD. I've ran Intel Burn In Test for 100 passes, and then failed on the 110th pass. So 150 is the minimum for IBT and 3 hours is the minimum for Prime95.

Another great way to see if your stable a little faster is to run Prime95 WHILE playing a relatively demanding game. Do this for about an hour and this can be a good short test to quickly test for stability. Either play a game or run another demanding benchmark like Uningine Heaven at the same time as Prime95. I run this specific test for 3 hours usually. And I sometimes fail after an hour.

I've had plenty of so called "stable" overclocks in the past when I wouldn't test enough. I've unknowingly told plenty of people misinformation regarding my "stable" overclocks. But sooner or later reality kicked in and I got a BSOD. Finally letting me know I had to up my voltage. All this has changed me now, as I now make sure the overclock is stable before I do anything on my CPU that I could possibly loose.

The worst thing about unstable overclocks isn't only BSOD's either. If a game your running uses 1-4 cores, and one of them becomes unstable it'll drop the core out. Causing obvious stutters in games and even in some programs. And the worst part is, you'll never notice that the frame rate drop was because of your overclock, so you'll leave it this way and keep having the same problem. Unstable overclocks also cause freezing on a computer. So with this overclock you have, you could very well be experiencing all of this and not even know it. It's kind of like when you overclock a graphics card too much and you start getting lower minimum FPS and experiencing stutters. Also check for WHEA errors as this is a sign of an unstable overclock.

All of this is why it's a good thing to thoroughly test your CPU overclock BEFORE you go on to your normal day-to-day PC activity. And won't you feel a lot better about your CPU knowing it can handle anything you throw at it and not BSOD or drop cores?