Monster QX6850 on Vapo LS

[Charles Wirth]

Specs:

Intel QX6850 ES
Abit IP35 Pro B04 bios
Corsair PC2-10000 Domaintor
Enermax Galaxay 1000
Nvidia 8800 Ultra
Vapochill LS R507 Chilly1 modded
4x Raptor 150's RAID 0
2x Gigabyte I-Ram Raid 0

Sub cooling:

8K BTU portable ac unit blasting side of case and front of Vapo.

Notes:

All benches the memory was running 1:1 3-4-3-9-1T @ 2.4v
CPU voltage

1.575v reaches 4.5Ghz
1.65v reaches 4.6Ghz
1.7v reaches 4.7Ghz
1.8v + didnt help with the exception of the top speed screen shot.

Benching with more than 1.7v on the vapo brings it down, even with the extra sub cooling, short runs like pi, 2k1, and AM3 were possible at 4.7Ghz

Pretty much everything else is 4.6Ghz

I did have a blow out of the raid drives where I lost a few good 3D scores but some I did upload and landed #3 in hall of fame and I used a backup drive for 2K1 and AM3.

The maximum I have taken a screen shot is 10x495FSB but I cannot upload that score. Here is the next best at 10x485FSB screen shot stable.


http://valid.x86-secret.com/show_oc?id=216637

First up, Aquamark 3 at 4.75Ghz, 700/1180


This took a while to get done and took 1.78v at this speed. At this speed only single threaded applications bound to the first core are stable.

http://service.futuremark.com/compare?2k1=9229412


Backing off to 4.6Ghz I am able to get off quad core benches like Cinebench.



Not sure how I did not get a super pi screen cap above 4.7Ghz but here is one at 4.65Ghz, I was climbing up and got distracted with 2K1.



One of the last things I was working on before the raid bit it was PCMark on the I-Rams in RAID 0. I uploaded an image that was fairly impressive at 485FSB...



One of the small problems was getting the single drive to boot with raid enabled in bios. I will replicate the benches lost and kick a little more ass in the process. I lost a few 3D runs with single card.

Right now I plan to rerun 3D with SLI on the 680i and currerntly reloaded on Vista again to play World in Conflict and finish DX 10 stuff now that I have more horsepower.

I was able to boot to 4.75Ghz and I did get into windows at 4.8Ghz a few times. Here is a screen shot of the general tag in control panels.



-----------------------------------------------------------------
680i and SLI results

Current #1 on the ORB "Vista"
http://service.futuremark.com/compare?3dm06=2651999


http://service.futuremark.com/compare?3dm05=3316849


http://service.futuremark.com/compare?2k3=5468938


-----------------------------------------------------------------------

P5K3-D results



4.7Ghz 760/1250

[CCUABIDExORxDIE]

VERY IMPRESSIVE! i cant wait to see what some of these chips do on dice and LN2

[Gautam]

Good lord Fugger, seems like you were hiding something in your back pocket all along. Those clocks are just nonsensically high.

[Exedy]

24/7 gaming with this "monster" must be something unreal

[Circaflex]

how did the raid blowout happen? scares me a little

[Charles Wirth]

Yeah, I held back. The 4850 should have been a clue.

I do have a 4950 that I cannot upload

[afireinside]

475*10 1t on a quad for 01... good god! nvm was thinking this was on the evga board. The abit P35 does look very fast though.

[Stapler]

^Sham ran his at 5ghz for '06. Not that hasn't been done with QX6700s, but I imagine it's easier to control temps.

Nice benches Fugger, that Vapo's holding up well.

[Charles Wirth]

Blowing out the OS usually happens when you take the ram too far out of spec. I believe I tried to boot at 495FSB 1:1 3-4-3-9 1T @ 2.5v

[[XC] gomeler]

Wow nice benchies especially for a modded vapo LS, bet the evap temps were dipping into the positives. Nice chip, plan on putting it under something a bit stronger?

[PanteraGSTK]

Nice on a vapo.

[Micutzu]

Very nice chip, what week ?

[Charles Wirth]

SLI next, I have a pair of Ultras begging for attention. I have some GPU cooling inbound as well.

Micutzu, not sure. Its an ES under the evap atm.

[Planet]

WWhens it going under ln2?

[Charles Wirth]

Cascade next, should hit/bench at 5Ghz I hope. Id even be happy with 5.1Ghz.

[njkid32]

Prolly not under LN2 but I'm sure we'll see some triple stage cascade runs

Great clocks FUGGER

[dinos22]

Blowing out the OS usually happens when you take the ram too far out of spec. I believe I tried to boot at 495FSB 1:1 3-4-3-9 1T @ 2.5v

could have saved yourself a bit of time and checked it quickly with a quick loop#5 memtest

[njkid32]

lol, FUGGER has a stack of hd's with installed OS's so all he's got to do is switch hd's and he'll be up and running in no time

[dinos22]

lol, FUGGER has a stack of hd's with installed OS's so all he's got to do is switch hd's and he'll be up and running in no time

my method's still quicker

too envious right now to congratulate you charges............these clocks are bloody insane

[Charles Wirth]

No it kinda sucked, I was benching on my RAID.

I switched back to Vista, new drivers and stuff released so I am giving that a try again.

[Planet]

You see shaminos on the EVGA 680i with P30 bios running 500FSB with the QX6850 and 5ghz 3d stable with ln2

[Emerica]

Wow congrats on the results Fugger, is that the same chip in your sig? Man G0 is beastlike.

[Charles Wirth]

Fredyama-san pointed out Kaz-n testing here;

http://blue.ap.teacup.com/kaz622/38.html



I didnt kick down in cores yet, I will see what gives.

I was setting up in the 680i and P30 bios and I may have broke the CPU socket with pressure. Never got any benches done on the 680i, it was mounted under the cascade.

[pmp]

Just for comparison (Cascade unit built by piotres, ~ -98oC):

http://valid.x86-secret.com/show_oc.php?id=217826

Above was done on VCore=1.70V (MB without any mods).
For sure it's not my last step on this Next try on LN2 coming soon... probably

[kemo]

Fredyama-san pointed out Kaz-n testing here;

http://blue.ap.teacup.com/kaz622/38.html



I didnt kick down in cores yet, I will see what gives.

I was setting up in the 680i and P30 bios and I may have broke the CPU socket with pressure. Never got any benches done on the 680i, it was mounted under the cascade.

That screen own anything we have ever seen