6800U overclocking

[r3b0rN]

with phase change/cascade cooling it generally goes the uglier the better

[Bennah]

that looks sick! Some of macci's setups are mad but dont look like that

[charlie]

and to think......
if the whole rig was just in an airtight box, you wouldn't need all the proofing
Easy to do too... just build an acrylic box 2' x 2' x 2', set in down on TOP of the rig, and fill the box with helium, like as in from ballons/tank.... no air, no moisture, no frost

C

[STEvil]

shh, they might hear you charlie!

[althes]

sweet setup

[bh2k]

Yea, lard Dabit over at ice cold has done it with the airtight case that totally stops condensation. I think in the area you live in it might not be a bad idea to check it out. I believe it's here

[Cow2kie]

That would be cool to have.. except when I have to reset my bios a million times on my dfi board

So would you not need any of the seal string on the prommy head if you had it in an airtight case like that??? How did he get all of the air out of the box once it was put together?

[zakelwe]

Originally posted by STEvil
shh, they might hear you charlie!

with his high pitched squeaky helium voice

Regards


Andy

PS Looking forward to Lardarses results, must be mid afternoon overthere in the Far East by now ...

Regards

Andy

[Hallowed]

Andy, my level headed friend:

Did you also notice a few things which have caused the "consensus" to lean unjustifiably to the X800XT?

Drivers: WHQL approved 60.72's with no Trilinear optimizations against Beta 4.5's which feature said optimizations and inflated 03' scores.

Clocks: X800XT PE's are 520 / 560, normal XT's are 500 / 1000, PE's are of a very limited quantity at over $500 pricepoint.

I think for gaming OR 3DMark, the cards are out and out equals.

[STEvil]

I'd like either right now.. 9700 died and i'm about flat broke :/

[zakelwe]

Originally posted by Hallowed
Andy, my level headed friend:

Did you also notice a few things which have caused the "consensus" to lean unjustifiably to the X800XT?

Drivers: WHQL approved 60.72's with no Trilinear optimizations against Beta 4.5's which feature said optimizations and inflated 03' scores.

Clocks: X800XT PE's are 520 / 560, normal XT's are 500 / 1000, PE's are of a very limited quantity at over $500 pricepoint.

I think for gaming OR 3DMark, the cards are out and out equals.

It would be a turn around if ATi are optimising and nvidia have stopped. There has been a 20% rise in some games in 3dmark03 over Cat 4.3's, so that is a pretty big jump. The 61 series for nvidia get 2-3% over 60.72, which is reasonable from I guess ongoing work.

I guess both parties really really want that #1 spot

I'm still awaiting my 6800U and it looks like it will be after Computex now.

Regards

Andy

[LardArse]

Ok I scrapped the crap zip tie F clampage and used some brass rods and screw to hold down better. I just can't seem to find screws this small around I'm sure clamping can be improved.






First run at 600/1190 with P4EE @ 323 x 14 -> 4522 :





Can't publish



I can't get any higher than 1.67v, maybe some overvoltage crap or something. Something's weird too, After a short benching, it keeps getting problems dropping back to desktop after a bench, I've had tests completed at 610 but black screens whenever it drops to desktop generating the results. Could be some condensation issues... Card also does not run as cold as I expected, the heatsinks and caps at the power area gets pretty hot with Evap @ -98C.

Nature seems kinda low for some reason...

Just the normal tweaks in nvidia panel. LOD seem kinda sharp, any ways to blur it? Trilinear optimizations on, which is faster?

[sierra_bound]

Nice score, Shamino. Nature does look low for those clocks. If you can get memory above 1200, you probably will have a shot at 17K.

[Hallowed]

Nice 1 LA.

That 6800U is packing some punch at those speeds.

[Crankster]

Aye, i smell a REAL battle! The first in a looong time.



PS. this ain't no prediction, it's a fact. Capische?

[Hallowed]

Capice, yes.

[charlie]

Originally posted by Cow2kie
That would be cool to have.. except when I have to reset my bios a million times on my dfi board

So would you not need any of the seal string on the prommy head if you had it in an airtight case like that??? How did he get all of the air out of the box once it was put together?

no, no, nooooooo
You don't NEED to seal the box airtight! make an open top acrylic box, turn it upside down on top of a mobo/rig.... fill with helium, which will push OUT all the AIR, leaving the mobo surrounded by helium, which DOESN'T have water vapor in it

C

[STEvil]

long wire + 2 way switch on end extending outside of box would for for bios resets.

I find it easier to flip PSU switch, wait 5 seconds, hold down insert, and power on.

[Kanavit]

hmm, what did Macci get with his top 3dmark03 score?

[ant1]

16774 :o
only 41 points more for a new wr!

[althes]

war i tell ya.

[charlie]

hmmm... so so I understand then that P4EE + LN2 + X800XT is basically EVEN with P4EE + Cascade + 6800U???
Wow....Nvidia's in the game, ehh?

C

[Cow2kie]

Originally posted by charlie
no, no, nooooooo
You don't NEED to seal the box airtight! make an open top acrylic box, turn it upside down on top of a mobo/rig.... fill with helium, which will push OUT all the AIR, leaving the mobo surrounded by helium, which DOESN'T have water vapor in it

C

no i didnt mean to seal string the box. I was asking if you would have to have sealstring on the prommyhead if you had it airtight like that filled w/ helium

[Kanavit]

Macci's x800 was at 770mhz.

Lardarse 6800 was at 600mhz. i can see why Macci is concerned about the 6800 ultra. He had every reason to be ..

[LardArse]

Charlie, I think the 6800 should go at least 20 Mhz more if I can get things goin, it needs more voltage, if only I could give it more, I think a 650 run with 1.8v and -98C evap is possible if I can get pass the voltage limit lock of 1.67v. Thing with nvidia cards is the damn throtling especially with memory overclock and you dunno if it clocks back.
WR right now is 1680+ with macci's Dual Cascades.