i7 920 overclocking thread

[gooddog]

Originally Posted by HardwarzFan

I think what the problem may be is the fact that people with 965 are thinking that their chips should be able to OC vastly higher than the 920's and 940's... but I don't think that this is the case. From what I hear the 965's can reach max OC's easier and at lower voltages, but I think that the entire i7 family has about the same roof. So if you are really wanting to get nutzo overclocks you may be forced to look more to cooling than chips.

I am not sure that is the issue. I agree with the first part, that they OC the same but the 965 requires slightly lower voltages.

However, to go from 4.0 to 4.2 gHz I am pumping in serious voltage to the chip and prime95 gives out relatively early but still doesn't hit serious temps. So I don't think it is a cooling problem. I think there is something else going on around 4G.

[MikersSU]

From reading other forums, it would seem that the OC wall is dependent on the processors. I'm seeing that the "B" batches are achieving higher clocks with lower voltages than the "A" batches. I wonder if this observation holds water.

[HardwarzFan]

Originally Posted by gooddog

I am not sure that is the issue. I agree with the first part, that they OC the same but the 965 requires slightly lower voltages.

However, to go from 4.0 to 4.2 gHz I am pumping in serious voltage to the chip and prime95 gives out relatively early but still doesn't hit serious temps. So I don't think it is a cooling problem. I think there is something else going on around 4G.

What about your RAM? Is it a high quality RAM, are you sure it mighht not be holding the OC back? Because I suppose if you can rule out the RAM, cooling, and chip it must be the mobo. Or if the motherboard checks out (people have had higher OC's then you on the same board) then it may be the chip.

[TFL Replica]

Originally Posted by MikersSU

From reading other forums, it would seem that the OC wall is dependent on the processors. I'm seeing that the "B" batches are achieving higher clocks with lower voltages than the "A" batches. I wonder if this observation holds water.

The opposite (A batches doing much better than B batches) has also occurred. So it really isn't a general rule that can be relied on.

[eflyguy]

My 4.2 "somewhat stable" (just need time to test it thoroughly) numbers are:
27x157, Uncore @ 3147. Vcore is only 1.375.
I've never seen temps over 70C..
..a

[SolShade]

Originally Posted by gooddog

I must be doing something wrong. I would hope that the 965 OCs better than a 920 - many of which seem to also get to 4.0 with similar voltages.

Ah yes i remember when i thought the extreme actually meant something when i got my QX9650. But unless you have a chip from the center of the wafer it's really not going to OC any better than the cheaper chips, or you can go LN2 and crank the voltage

[gooddog]

Originally Posted by HardwarzFan

What about your RAM? Is it a high quality RAM, are you sure it mighht not be holding the OC back? Because I suppose if you can rule out the RAM, cooling, and chip it must be the mobo. Or if the motherboard checks out (people have had higher OC's then you on the same board) then it may be the chip.

I have very good RAM (Mushkin, 1600) and have also tested the OC by running it at 1333 rather than the rated 1600, so I don't think that is the issue.

I get stable at 25x160 but going to 26x160 is not stable for 2 minutes under prime95.

[Fresh2death]

Just got another stable over clock of 3.8MHz this time with (HT) turned on. I turned off turbo boost mode for it to be more stable seems like 100% stable now solid as a rock to me.

here are my settings in case any of you would be wondering..

CPU Vcore - 1.31875 (couldn't get much lower then this with turbo boost turned off and with HT turned on)
QPI/vtt Voltage - 1.220v
IOH core - 1.100v
DRAM voltage - 1.500v
CPU PLL - 1.840v
==================
QPI Link Speed x36
Uncore Frequency x13=2470MHz
Load-Line calibration - enabled
==================
intel turbo boost tech - disabled
CPU enhanced Halt(C1E) - disabled
C3/C6/C7 state support - disabled
CPU thermal monitor - disabled
CPU Eist function - disabled
Virtualization tech - disabled
Bu directional PROCHOT - disabled

[Trigunflame]

*cry* definately going to be messing with my true soon... can only get 1 of my cores to idle @ 32c or lower atm, the rest are 3-5c higher.

[Fresh2death]

Originally Posted by Trigunflame

*cry* definately going to be messing with my true soon... can only get 1 of my cores to idle @ 32c or lower atm, the rest are 3-5c higher.

That's still not too bad, just when you go to re-seat it just make sure you have that thermal paste on so thin you can almost see threw it. That's what i did and I have great temps with my TRUE..