[Ch0p5t1ckB01]

Hi all!
Long time reader, occasional poster.
Recently picked up an 920 d0 (specs in the sig) to replace my "aging" Q6600.
The first thing I... does anyone wanna take a guess?
... Ok, enough with the bad jokes... I'll skip to the good/bad part.
Overclocked it to 3.6Ghz @ 1.264 vCore without any issues... 3.7 required a bump to 1.298, and 3.8 @ 1.312. All have passed Prime95 and have been stable. So I tried my luck and went for 3.9 Ghz. My first pass of P95 crashed 30 minutes in. I didn't want to increase my vCore as it it already @ 1.337 in the bios (1.312 in Windows under load), so I lowered my UCLCK Multi to 17x instead of 18x, and currently it has been P95 stable for 50 mins. I feel that I picked up a bad batch/chip because most d0 owners can get 4Ghz with 1.20-1.25 vCore, not to mention that my temps are 67-73 depending on the core. Thoughts?

The following are my settings in bios... maybe some of the veterans OCers can help

Cheers!!

CPU ratio = 20
BCLK frequency = 195
PCIE frequency = 100
DRAM Multi = 6x
DRAM frequency = 1171MHz
DRAM timings = 7-7-7-20
UCLK Multi = 17x
UCLK frequency = 3315MHz
QPI Multi = 36x
QPI link data rate = 7020MT/s
CPU voltage = 1.33750V (Bios)
CPU PLL voltage = 1.88V
QPI/DRAM core voltage = 1.35V (I think, I'll have to double check)
IOH voltage = 1.2V
ICH voltage = 1.2V
DRAM bus voltage = 1.64V
Load line calibration = enabled
Intel HT = Disabled
Intel Speedstep = Disabled
Execute disable bit = enabled

 

[mothow]

Not bad.And a push/pull Corsair Hydro H50 nice

 

[Stellarex]

Seems the newer D0's are starting to do this for some reason..

 

[Ch0p5t1ckB01]

I forgot to mention that my batch is 3913A909. Anyone else with this batch?

 

[JonnyMark]

Wish I can help you out with you temps but you have HT disabled, but on the note people getting 4.0GHz on 1.2-1.25, that's not all ways true, I needed 1.26250v to get 4.0GHz.

Do you feel like you overclocked too fast?

Also search your batch d0 batch number I read a post little while back where people posted their batches and overclocks.

Also you still got room, max temps are 85C, max voltage is 1.35.

Also to be honest little noticeable performance from 3.6-4.0, real world wise.

 

[ljason8eg]

I've found the 21x multi to be much more stable than the 20x multi. I need less vcore for the same clock speeds, and 200x20 seemed impossible to get stable for whatever reason, but 21x191 was extremely easy.

Just some stuff to try if you want to shoot for 4.0

EDIT: Oh HT is disabled? Ouch on the temps.

 

[Ch0p5t1ckB01]

Quote:

Originally Posted by JonnyMark Wish I can help you out with you temps but you have HT disabled, but on the note people getting 4.0GHz on 1.2-1.25, that's not all ways true, I needed 1.26250v to get 4.0GHz. Do you feel like you overclocked too fast? Also search your batch d0 batch number I read a post little while back where people posted their batches and overclocks. Also you still got room, max temps are 85C, max voltage is 1.35. Also to be honest little noticeable performance from 3.6-4.0, real world wise.

1.26250v is considerably less than my 1.312

As for overclocking too fast, I don't think so... based on the results from most users, 3.6 is easily obtainable. Besides I've done it before with my Q6600, and Opteron 165.

There's a post @ XS with Batch Numbers... sadly there isn't a lot of 3913s.
I'm sure I won't notice any difference between 3.6 to 4.0, it's just that I'm aiming for 4

 

[Ch0p5t1ckB01]

Quote:

Originally Posted by ljason8eg EDIT: Oh HT is disabled? Ouch on the temps.

Tell me about it

 

[ljason8eg]

Dunno how much it would help, but better fans on the rad would help drop temps at least a bit.

 

[Ch0p5t1ckB01]

Quote:

Originally Posted by ljason8eg Dunno how much it would help, but better fans on the rad would help drop temps at least a bit.

The Scythe Slipstreams are rated for 68CFM @ 24 decibals... but I'm sure that's not the actual rating.

I just didn't want to have a jet engine in my room