Overclock "diary" for my Intel DP67DE

[frontier204]

EDIT: Was thread-jacked by a new post, but is relevant to the topic of OCing this motherboard: Overclock "diary" for my Intel DP67DE

Hi all,

Update: I achieved a just barely stable 4.0GHz from my Intel DP67DE (one voltage tick down and Prime95 fails)

The latest BIOS for this board allows some actual overclocking attempts without the BIOS going nutty and putting dumb RAM settings in for you which refuse to post. The OCing bottleneck for this board is the "naked" MOSFETs that provide voltage for the CPU. If they get too hot, BAM you're down to the non-turbo multiplier of x33.
The highest OC I managed with a Megahalems was around a x38 multiplier, and even that would downclock if you run extreme heat stress tests like IBT / Linpack. A downdraft cooler helps a lot, but it seems this board still maxes out at clock speeds that other motherboards could give you with their "auto overclock" setting.

[frontier204]

As an update, I changed my BIOS to version 0064 released early this month, and finally allows me to OC without a pile of no-POST.

I know my chip is stable at 4.3 GHz without much of a problem, but my motherboard will only allow me to reach 3.6 GHz before throttling the CPU.

It seems any higher and the CPU drops back down to the stock 3.3 GHz because Turbo mode shuts off. Note I cannot adjust my non-turbo multiplier.

[frontier204]

Second update - I managed to hit a 3.7 GHz without getting throttled... by switching from my Megahalems to the stock Intel HSF.
Despite the CPU temperature hitting 10 degrees Celsius higher than what I got on the Megahalems (according to Intel Desktop Utilities), there is no throttling. Yes I tried re-seating the Megahalems changing how I used intake and exhaust on the case.

That makes me believe that the crucial difference is the fact that the Intel HSF is a downdraft cooler, which is cooling the uncovered VRMs. With the Megahalems, there was no airflow in that region because the mounting bracket actually covered the area where the VRMs were. There was something like 2-3mm between the Megahalems bracket and the top of the MOSFET, so only an ear-piercing loud fan could've forced air fast enough through that area to actually cool it. By comparison, the Intel stock HSF's incidental airflow is obviously hitting that area. Maybe with a better downdraft cooler that still has a smallish (not bigger than the socket) base I can OC this poor Intel board more?

I wish I knew this during the Chimps Challenge - I would've gladly sacrificed fan noise for the extra points, however small the bonus might've been.

[gingerbee]

having some wierd problems myself just wondering if its me not use to efi bios

[frontier204]

I didn't really notice anything special when I flipped the UEFI mode on and off on my board - it's hard to pin down a specific change as being at fault for something with all of the errors being inconsistent. The BIOS for my Intel board is quite obviously bugged - nothing like my old ASUS AMD build that just worked.
I just came back to find my computer hard-frozen and an error message that said vPCH was out of bounds...

[frontier204]

Well I managed to squeeze another 200MHz OC out of this board. I set a really low VID so the motherboard's habit of applying way more voltage than it actually needs in turbo mode is lessened.

I had to set the voltage control to "static" (which doesn't actually mean static if turbo core is on) and kept the amps and watts settings for the turbo core limits at their defaults. Otherwise, the board will downclock to non-turbo after a while. With a VID of 1.1V and the all the turbo multipliers set to x38, I get this:


I can't push it much higher than x38 multi because of the downclock limit. Even at x38 it will downclock to x37 running Prime95 or so because of the board exceeding the turbo mode power limit. Anyway it doesn't downclock when doing SMP folding so that's what matters for me

BTW, this is back with my Megahalems that has a mounting system that chokes the VRMs from any form of airflow, so this isn't too bad.

[frontier204]

When I played musical cases with my rigs ( Server -> Lancool PC-K62, Main -> Antec Mini P180), I changed the Megahalems in my computer for a Cooler Master GeminII.
With a 140mm fan going at full speed, I was able to squeeze another 200MHz out of the motherboard.

So I'm now running a x40 multi (4.0 GHz) for anything other than Linpack (in which case the VRMs try to save themselves from smoking and downclock the CPU to x37). At least I'm able to use the capability of my i5 2500K now, as the non-K would have topped off at x38.

[gingerbee]

glad to hear it i was just not use to the UEFI bios and now i am getting great overclocks on all my boards glad you got yours to somewhere your happy with

[frontier204]

I don't think UEFI has any effect at least at the BIOS revision I'm using, but I keep it off anyway as my biggest boot drive is 250GB (Linux install).
I still wanted to see the 4.2 GHz that I got my old Gigabyte P67 UD3P board to before I had to send it back for the Cougar Point B2 bug, but hopefully the OC to 4 GHz will hold its own for games and my school tasks for 2 CPU generations (which is my next planned CPU/Mobo/RAM upgrade).

[deadlyscar]

i'm new to overclock but i understand most of the stuff can i have a simple guide to overclock my 2400 with my intel DP67DE board plz .