x7337H

So after hours of testing this and that, I believe I have come to the horrifying realization that I have hit not an FSB wall, but a plain and simple Ghz wall (if there is such a thing).It's like this: I currently have my proc running at a stable 3.4 (400x8). It's cool, voltages are a little higher than what I'd hoped, but it runs. Recently though, I've tried to get some
more performance going. I found out that my RAM (though maybe my motherboard is limiting this, reviews say otherwise) can get to 470Mhz with a
x6 multi. Orthos failed @ 475x6. Regardless, I simply wanted to get a stable 3.6 with my Conroe, and have tried numerous configurations.
Orthos fails immediately with 400x9, 450x8, 455x8, and 465x8. I thought this might be an FSB hole, is this the case? My ram, according to a 3rd party review, should be able to hit close to 500Mhz on nigh stock voltages and 5-5-5-15 timings. Currently, for fear of instability, I'm running it @ 2.2v (+.1 over max stock, as "stock" for the RAM was reported to be "1.9-2.1"). I'm running my proc @ 1.424 load. Now NB, SB, and VTT are a whole different story...My mobo, not being particularly OC friendly, has simply assigned the voltages for NB, SB, and VTT the qualitative values of "Lowest, Low, (something along the lines of) Medium, High, and Highest." This is not the case for CPU voltage or RAM voltage, they have specific numbers (the least ASRock could do).I have disabled all options in my BIOS that might hurt any OC results, but I am unsure about three things: CIR10 Field, DRAM RCOMP Setting, and (Memory) Flexibility Option. CIR10 Field 1(I think there's a one after it) can be set to Auto, or disabled. When I disabled it, I found that instead of Orthos simply telling my rig was unstable, Orthos shut down my rig altogether. I've kept it set to Auto. DRAM RCOMP Setting seems really complicated. I currently have it set to Auto as well. THe other option, Manual, allows you to set up some 15 or so part combination with each part consisting of 2 letters and 2 numbers. I have absolutely no clue what any of the DRAM RCOMP settings are. Last comes the (Memory) Flexibility Option. This can be set to either Enabled or Disabled. Enabled, it claims to "increase memory compatibility." I've found, both in personal experience and from another user, that this "increases compatibility" by downclocking your RAM by 5Mhz. I really don't know if it increases any type of compatibility, and I forget if it's on or off right now, but if it is on, I've simply raised my bus speed/ RAM speed by 5(Mhz). I would like to think that it's simply my processor and/or RAM that's limiting my OC. I'm looking into an E8500 and some Corsair DDR3-1600. COuld My E6600 (B2) simply be a poor overclocker? I've seen OC results on my board. They can get E8400s to 4.2, albeit they've tweaked the divider, which I'm having a little trouble grasping, but more trouble changing. Part of this may be due to the fact that my board, despite its new X48 design, still uses jumpers to switch the FSB/FSB (or is it NB?) Strap. Currently, I have the jumpers set to accept 1600Mhz FSB, before I believe it was set to Auto, or maybe just 1066Mhz. Regardless, this might be a problem. My rig has buckets of peculiarities, and I won't list them all. What is holding my overclock back?


Sorry about the blinding wall of text...and sorry if the format is weird, I made the mistake of writing this in notepad.

The Duke

Yes, persay there is!
There was an indepth how and why in one of the DFI Dark MB posts. I think it was the P35, 38 or 45....
Where as an issue with voltages and the effects can likely be countered with some serious tweaking.
I'll see if I can find the link.
EDIT: Sorry, I can't seem to find the clunk.org link to the write up on the relative info.
In a DFI MB, there is an additional voltage tweak that in effect ballances voltage to a very fine degree. My understanding is various harmonics will cause a flat spot in the OCing where as some try to jump past that speed in hopes of being able to OC byond it. Where as the additional BIOS attribute in the DFI MBs can cure it.

__________________

subliminally incorrect

AFAIK Asrock brand motherboards is the budget level motherboard series released by asus, and by the looks of it they can allow you to overclock.

Isnt your stock multiplier at 9? (you say you got 3.4GHz (400x8 is 3.2GHz))

Take my rig into consideration here: ive got an E6400 and ive gotten it from 2.13GHz stock and ive pushed it to 3.4GHz and I run into a similar problem as you, no matter what adjustments i do i cant go anyhigher. I thought it would be a ram problem as i had a pair of corsair select value RAM and decided to purchase some mushkin ram that can run at 533MHz FSB. popped them in and tried to go beyond 3.4GHz (425X8) and still no avail.

What im trying to get at is i think you've reached the limit for your processor.

__________________

x7337H

thanks buckets Duke. So you really think that I could have hit a wall, eh?

@Subliminallyincorrect: Yes, stock multi is x9...could an E7200 (leaning towards that now) fix it?

The Duke

Each CPU has its limits. Those are often compounded by the MBs capabilities and ability to control vdroop and perhaps lack of BIOS features to combat the OC.

__________________

x7337H

Theoretically, what's a good Conroe Dual Core OC? In my opinion, though 3.4 is a peachy OC, it's a little low for a C2D.

I believe it was a HardwareCanucks review, but they said that they had some trouble getting the board stable with certain dividers. I'll try to find that review...

Takendown2

mine wouldnt go past 3.52 no matter what volts i give it, i even tried 1.5Vcore with 1.55on the NB but it still didnt work, its the actual limit of the CPU and also ur MB is not very good for ocing, i read somewhere that AS Rock Boards cant even get close to a FSB of 500 which other board eg ASUS,DFI,Gigabyte board can do easily

__________________
Sorry for my bad english I am not a native english speaker.
NO SYSTEM IS EVER GOING TO BE FUTURE PROOF!!!
3.36 GHz on E6420 1.35Vcore:http://valid.x86-secret.com/show_oc.php?id=394404
4GHz on E7300 with 1.4Vcore(1.3625V After Vdroop):http://valid.x86-secret.com/show_oc.php?id=443542
Refer To My Sig Rigs for Different Builds!(Click on them and you'll know what i mean)

x7337H

Here is the HWC review concerning the overclocking results, and the divider problems.

Does this shed any light on my situation?

x7337H

bump

Are my dividers the problem?

I'm thinking about taking my Xeon X3220 from my brother's rig and seeing if I can get a higher clock.

Should the quad OC better than the Dual? I know it really shouldn't, but could I just have a bad E6600?

oneacn

I just updated my sig rig, it is current and complete I wish I could give the duke rep but for some reason I can not. He built my sig rig for me. I have had no problems with it at all, the board while rather old can still be found online, it is an AMD SLI 939 board that is very overclock friendly it is advertised as being overclock friendly. I found one somewhere I forget where for $220.00 If i needed a second rig like my first I would jump all over it and buy it again and do the exact same setup for another computer in my house.

The board is the DFI N4 SLI DR I run my 1.8 Operton 165 @ 2.7 stable, I could crank it to 2.8 or 2.9 with my case I have but not really wanting to push out that much of the system for just playing WoW and Americas Army no need to.

So in conclusion if you can find the DFI Board I have jump all over it. Good Board

__________________



[center]

World of Warcraft Players club

[center]