X3350 Arrival

[trans am]

Hmm... after some measures, I've found that the true vNB required for 4GHz is... actually real 1.732v - 1.742v. A northbridge voltage mod for ASUS P5E-VM HDMI is possible... but seems like we'll be dealing with adding resistors instead of pencils... oh well. But my current board already has a +0.018v measure on the northbridge volt by default... so setting 1.71v in the BIOS gave me 1.732v real... barely enough for 4GHz... and that explains why I was able to get 4GHz to go more stable on this new board. More reports to come in later!

so much talk and no ss. lets see some pics

[Northwood]

are these multi unlocked?

[Kensek]

I had to skew Dram Clocks with in Bios.

Cha A + 150 ps
Cha B + 250 ps

Other Bios Settings

CPU v 1.3875
CPU PLL 1.60
NB 1.69
Dram 2.30
FSB Vtt 1.60
SB 1.175
SB 1.50 1.70
LoadLine Auto
Spread Specrums Disabled on CPU and PCIe


Lavalys shows a little different from Bios Settings




http://service.futuremark.com/result...&resultType=14

[Kensek]

Time to Drink Now - My Mission has been completed.

[LagunaX]

Time to Drink Now - My Mission has been completed.

[conzymaher]

What would people say is an "easy" overclock with the X3350? 3.4ghz?

It just doesn't seem as appealing anymore with Q6600s being a full 100 euro cheaper and reaching 3.6ghz easier and the E8400 reaching 4.3ghz easy...

Do any video converters take advantage of all 4 cores? I have tried Super, Any Video Converter, TMPGEnc, iriverter, AutoGK.... all only use maybe 70% of one core and 50% of the other on my dual core server...

So im more inclined to go with fast Dual core than a Quad that may not ever get passed 3.4ghz...

[Kensek]

3.40 is easy.

But yes, the new 45nm Quads are just barely fractionally better than the 65nm Quads.

[LagunaX]

This seems fairly easy:
http://www.xtremesystems.org/forums/...d.php?t=182499

[conzymaher]

How many people here are using air cooling? Think a true or a IFX-14 would get me to 3.6ghz with a q9450 / x3350?

[Kensek]

I use a Lapped Zalman CNPS9700NT. I'll never use Water!!!

[conzymaher]

hehe I was planning on water but its to damn expensive :P

Going with a IFX-14 or a True now I think

[sofarfrome]

I use a Lapped Zalman CNPS9700NT. I'll never use Water!!!

I used to say the same thing. But now that I went WC I will never go back. Not in my rigs at least.

[RunawayPrisoner]

so much talk and no ss. lets see some pics

Only one at the moment. :p



^= What I've found thus far. Sanding those resistors help... but will take forever...

[Kensek]

Sand Away those Resistors!!

[RunawayPrisoner]

Yes! I actually did (just a bit, though... an hour was too long for me...) and now my real vNB is at 1.735v.

[Toker5000]

Batch # L803B322, packed 03/26/08. Maximus extreme, Vcore 1.4 VPLL 1.63 (1.56 bios), VNB 1.57, increasing did not help (maybe cause of overheat?) FSBT 1.4V cpu gtl x63 (max) nb x67

Can't break 450x8 stable... posts ok till 480fsb, crashes shortly after starting any sort of testing program. Same wether im running 1400's memory or 1800mhz.

I might be having nb temp issues, im not watercooling my maximus extreme anymore..

[trans am]

do I have the highest oc on x3350? I haven't seen any screen over 490 fsb

[rx7speed]

I would be VERY surprised if it was your PSU holding you back. To find out if it is or not, get a cheap digital multi-meter and put the probes into a molex plug (preferably +12v and ground) while running 3DM06. If you see a big voltage drop just before it crashes (eg if the voltage drops to like 11.0v or lower), then it's your PSU, otherwise, it's most likely instability elsewhere.

Eller

could be noise on the PSU though that might cause some instability. so even if he still has over 12v the rippled might not help very much and that might not be detectable with a standard DVM

[RunawayPrisoner]

do I have the highest oc on x3350? I haven't seen any screen over 490 fsb

Well, 495 was bootable, and I wouldn't have been able to validate 500 if I couldn't get into Windows. 490 was the max benchable, though. 3DMark06 score:

[FiXeR]

I've a Xeon X3360, but I think I have a FSB wall at 444Mhz, what is incredible low. What are your BIOS settings? I think that the CPU can get higher.

[trans am]

Well, 495 was bootable, and I wouldn't have been able to validate 500 if I couldn't get into Windows. 490 was the max benchable, though. 3DMark06 score:

Thats what I thought.

[trans am]

Well, 495 was bootable, and I wouldn't have been able to validate 500 if I couldn't get into Windows. 490 was the max benchable, though. 3DMark06 score:

Thats what I thought. all talk

[RunawayPrisoner]

Thats what I thought. all talk

I know... and I apologize for the lack of pictures. There isn't anything useful anyway... as this Raidmax PSU that I have at the moment won't let me stay stable at any settings other than stock. In two days, maybe I'll have shots of completely stable overclocks on the Q9450 soon. Or actually, wait a sec, I'll see what's the max clocks I can boot into Windows with right now. :p (I can at least take shots from Windows)

[RunawayPrisoner]

BUMP! And FYI!



I believe that voltage would be the stable volt for 4GHz... but I can't confirm it until I have some better stuffs two days from now...

Edit: And the graphics adapter doesn't show in the validation for some reason. D:

Grrr... onboard GPU doesn't show... But you should know that it's there... hence Vista Aero (look at the window of CPU-Z)

Anyway, 505 FSB! Seems like the northbridge volt mod... works!



And... I wonder if I should...



Edit again: OCCT crashed after 2 seconds into the stress test. And here's the validate link:

http://valid.x86-secret.com/show_oc.php?id=345603

[zoob]

Finally received ONE of the two I ordered:

L803A756

Installing it soon..