• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

OC log: Penryn Quad on GA-X38-DS4

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Albuquerque

Member
Joined
Mar 26, 2003
Location
North America
I'm starting this thread because it is driving me NUTS that seemingly nobody has overclocked a Penryn quad on anything better than a P35/G35 chipset. Why spend $350 on a processor if you're gonna stuff it on a $89 motherboard? ;)

(no offense intended to those who are stuffing said $300+ processors on said $89 motherboards ;) )

Anyway, much props to the Tank Guys for having the Q9450 in stock, I ordered one ten minutes ago and this thread will be my log of exactly how it does on my X38-DS4 under an Apogee GT and dual 120mm rad.

To give you some background on my motherboard: I've had it running 575FSB with entirely stock voltages across the board on an E6850, with exception to the DDR voltage as my poor Ballistix DDR2-800 needed some help to hit 1150 stabily ;)

So, I know the board is more than capable of hitting uber FSB's, so the only drawback now should be the processor itself.

TankGuys Sales said:
Order Number: xxxx
Detailed Invoice: https://www.tankguys.com/account_history_info.php?order_id=xxxx
Date Ordered: Thursday 27 March, 2008

Hello from Albuquerque @ OCForums.com :)

Products
------------------------------------------------------
1 x Core 2 Quad Q9450 2.66 Ghz (Yorkfield) TRAY/OEM (EU80569PJ067N) = $359.99
------------------------------------------------------
Federal Express (1 x 1lbs) (Home Delivery (2 days)): $11.05
Sub-Total: $359.99
Total: $371.04

Let the saga begin! :beer:

Edit
Shiggy requested that I pop into this first post and put up some baseline information so it's easier to find later in life when the Q9450 is "old news" ;) So...

Q9450 SLAWR 2.66Ghz/12Mb/1333
FPO:L803A55 Batch: 357357

I never tested the temps stock, but at 3.2Ghz (400 x 8) on stock volts the two working thermal sensors indicated 44* after about 45 minutes of 2 x Orthos with 10k in-place FFT's at priority 10.
 
Last edited:
I was looking at this board, have you tried a 45nm dual-core CPU on it? Whats the highest FSB you've been able to obtain? Considered flashing it up to the X48 model?
 
You read my sig? ;)

My E8400 tops out around 535FSB give or take, I could probably squeeze more out of it if I really wanted to get evil, but with a 9x multiplier I really don't need that much anyway. Still, I do start hitting an FSB wall around 466 that necessitates a bump in vTT of 0.010v -- 500FSB needs a vTT bump of 0.020v, and >525 needs a vTT bump of 0.030v. I have more options after that, and I'm still on stock MCH, so I assume I could continue pushing it until I get tired or burn something up.

The sweet spot on this board for memory performance is right around 482, and that's what I run for my 24/7 overclock (4,338Mhz).
 
on an E6850

Unfortunately dual-core FSB is not equavalent to quad-core FSB, regardless of architecture. Hopefully these new quads show us something and you've got yourself a nice combo. 4GHz would be a win in my mind, with anything beyond a bonus.
 
Unfortunately dual-core FSB is not equavalent to quad-core FSB, regardless of architecture. Hopefully these new quads show us something and you've got yourself a nice combo. 4GHz would be a win in my mind, with anything beyond a bonus.
Oh yes, absolutely. If you read my post above yours, the FSB difference between E6850 and E8400 was pretty noticeable too.

Still, I know this board has considerbly more in it than most P35 chipsets -- how many P35's can claim 575FSB on stock everything no matter what processor? I haven't seen one yet.

I mentioned that, because of this: If I run into an FSB wall, it's quite assuredly something specifc to the processor rather than the board. The scant few attempts at overclocking X3350's (here, over at XS, and a few other places that Google found) were all people on P35's or G35's that were having FSB problems even before they got the quad. That says something to me...
 
Any chance of lowing your E8400 multi to 6x, and just try cranking the FSB to see what the board top out at while making the proper NB and VTT voltage increases? That is of course after you get done testing your new quad :D
 
Any chance of lowing your E8400 multi to 6x, and just try cranking the FSB to see what the board top out at while making the proper NB and VTT voltage increases? That is of course after you get done testing your new quad :D

Both of the above FSB numbers were attained at the 6x multiplier for both chips. I could POST on the E8400 all the way to 560FSB, but it wasn't Orthos stable at any more than ~535FSB. I suppose I could get further if I really wanted to hammer on it, but it probably isn't worth it.

Chip ships today, so hopefully I should have it on my desk sometime Monday... Thanks and props again to TankGuys :beer:
 
Will be here tomorrow: :beer:
Ship date
Mar 28, 2008

Estimated delivery
Apr 1, 2008

Destination
Louisville, KY

Service type
Home Delivery-Direct Signature Required

Weight
1.0 lbs.

Won't have any results until after work (~5:30pm EST). Gonna see how far we can get it on stock volts first...
 
Yeah Im looking forward to it also since Im thinking of doing another build and giving this one to my wife.
 
Can't wait to see results. If these quads overclock better than what they are rumored to do, then i might have to picky myself up one :drool:
 
I too am excited to see what these things are capable of, even though I won't be able to get one myself (just bought a 3870 and waterblock to go with it).

Good luck on the OC!!

~jtjuska
 
Mine just came in and I have the same motherboard too. Hopefully I will find some time to play with this tonight.
 
First impressions: Board recognized it immediately as 1333FSB Core 2 Quad Q9450 -- VID is 1.2375 (running the F2 bios). I seem to have at least one entirely stuck thermal sensor, and another one that's iffy using RealTemp 2.24 to check.

Posted at 400FSB x 8 (3.2Ghz) on stock volts, and ran about 30 minutes of 2 x Orthos without issue with temps in the low-40's. Next, posted at 450FSB x 8 (3.6Ghz) on stock volts, but was not Orthos stable for more than about five minutes. I'll give it a bit of massage there and continue on...

Edit
False alarm, my Crucial Ballistix needs more than 1.8v to do DDR2-900 :bang head

Now Orthosing at 450 x 8 (3.6Ghz) on stock volts :beer: If 2 x Orthos is still running in another half hour, we'll go to 475...
 
Last edited:
Wow, there's a monsterous FSB wall at the transition from 455 to 456 FSB. It POSTs fine, but cannot start Windows PE for testing. A few quick adjustments to MCH and VTT didn't seem to do anything, so I'll have to continue fidgeting with it to see where the problem lies.
 
Wow, there's a monsterous FSB wall at the transition from 455 to 456 FSB. It POSTs fine, but cannot start Windows PE for testing. A few quick adjustments to MCH and VTT didn't seem to do anything, so I'll have to continue fidgeting with it to see where the problem lies.

Hope you can go at least a bit higher. Did you try any higher than 456? Could be a FSB hole, never know! But honestly, and sadly, its not unexpected given what we already seen from 45nm Quads. And im not so sure there will be a revision that will allow for higher FSBs, im afraid the next revision will move to FSB 400 with even lower multis, the QX versions will still be the only option for 4GHz +. :((
 
Back