SLI + P5N32-E results

[G H Z]

I decided to pick up some new toys recently and after a week or so playing around I finally have some results. This board is certainly not as easy to setup as the Abit AW9D-MAX but I'm finding out it's quirks and getting the settings down for some real benching. Unlinked on 680i is interesting. Apparently there are segmants in the dividers working behind the scenes as the actual memory speed is rarely what you type in. Performance wise I have yet to see a real big difference for 3D, in fact I have been running well over 550mhz CAS 5 and it's comparable to 420Mhz CAS 3 settings - still it's fun to experiment with it.

For now this is air testing and I think these are all pretty much the best I can get with this cooling. The next step is to freeze the processor and leave the cards on air and then finally ice everything down. The IQ with single/SLI is amazing, in fact I'd say better than any 'other' setup I have played with. Of course this is only in 3DMark, I have no idea about gaming

Setup

X6800
P5N32-E BIOS 602
BFG GTX's
Corsair 6400C3 Unlinked 570mhz 5-4-4-4
PCP&C 1KW-SR



01


03


05


06


Aqua - SLI in Aqua is rather boring, I've already managed 244K single GTX

[bazx]

these cards are very nice

the above runs are without vmod?

[eva2000]

nice... try 430-450mhz 4-4-4-x 1T much faster i.e. 434mhz 4-4-4-12 1T faster than upto 533-550mhz 4-4-4-12 2T even for 3dmark06!

[CHRONiK]

For now this is air testing and I think these are all pretty much the best I can get with this cooling. The next step is to freeze the processor and leave the cards on air and then finally ice everything down. The IQ with single/SLI is amazing, in fact I'd say better than any 'other' setup I have played with. Of course this is only in 3DMark, I have no idea about gaming

Hi GHZ,

How are you going with the northbridge chip heat and PWM heat others have been describing, have you needed to do any mods for this?

[Dynasty]

nice results for air.......

Is the P5N32-E vmodded ?

[TaPaKaH]

Nice results

At least 03 score (don't know about others) is a WR with "all-air" setup

[msgclb]

X6800 $958.00
P5N32-E BIOS 602 $309.99
BFG GTX's $609.99 ea ($1219.98)
Corsair 6400C3 Unlinked 570mhz 5-4-4-4 $479.00
PCP&C 1KW-SR $599.99

Total $3,566.96

Results - Priceless

** prices are from online retailers not necessarily in stock and probably don't reflect the true cost

[wa77ss]

Those results are in no way priceless.

[Gautam]

Super strong in 03 for sure. Looks like you're getting pretty CPU bound in Aqua, which will be no biggie in the future I'm sure. 01 is quick as always. Seems like only a couple of guys are quicker than you at sub-4 GHz.

[G H Z]

No mods on these runs at all guys, stock board & cards. NB is cooled by 80mm fan, no problems with NB overheating. It does run very hot without active cooling.

i see something nice in teh background? yet you prefer air?

I know, unfortunately that Baker unit won't mount on 775 yet

[PcCI2iminal]

congratz GHZ
love the 3d03/3d05/3d06 scores

[Brettbeck]

Nice scores !

LOL !

[MTP04]

Very nice results GHZ!

What vcore were you using?

[G H Z]

Thanks guys

I set 1.625 in BIOS, this board undervolts so it's really around 1.6 loaded, the max this chip can take on air. I found a nice kick in system power running tighter latencies as Eva suggested and thanks to hipro's excellent strap testing on Striker seen here. Running a little less mhz added around 500-1K to each result and amazingly 18K to aqua, fun stuff.

[dinos22]

Those results are in no way priceless.

LOL good observation 99



GHZ would be interesting to see what 1T would do to those scores as eva mentioned

[NiCKE^]

G H Z, you should thank Hicookie for the excellent strap testing

[Kunaak]

for me, 900 mhz, cas 3-3-3-4-1T and TRC of 1 to 4, is faster then 1300 mhz, cas 5-4-4-10 at 2T and TRC 10.

give that a shot.
I have the same board, but only use a single GTX myself.

[k|ngp|n]

Having fun yet G ?? Take your time and get to know everything before you go to ln2, then give those cards hell . Looking great so far

[G H Z]

You're right Kunaak, boot 424 then CG. Check out this air cooled Aqua

Having fun yet G ??

Yeah it's OK, well see how much fun it really is in the next couple of weeks

[eva2000]

GHZ how much vdimm you use for 445mhz 3-3-3-8 1T ?

[Gautam]

You're right Kunaak, boot 424 then CG. Check out this air cooled Aqua




Yeah it's OK, well see how much fun it really is in the next couple of weeks

I can't believe my eyes...it seems like in your hands 680i could very dethrone i975X as the performance king. 445 CAS 3 1T is just nuts.

[Kunaak]

You're right Kunaak, boot 424 then CG. Check out this air cooled Aqua

wow, someone actually took my advice.
dont see that too much these days.
seems hard to get anyone to believe that 900 mhz at 1T can be faster then 1300 mhz at 2T, but its definatly true in some cases.

alittle advice.
you dont really need Tras 3.
sacrifice alittle tras, and get your TRC as low as you can go.

I can get Cas 3-3-3-4-1T, and Trc 1, at 900 mhz, with some plain jain D9GMH, and you definatly got some good ram too, so you should be able to get alittle lower then 11.
believe me, it's definatly worth it, when going for low latency.
Trc is a hard one to push, but if you can get lower, its definatly worth it.

[MTP04]

Thanks guys

I set 1.625 in BIOS, this board undervolts so it's really around 1.6 loaded, the max this chip can take on air. I found a nice kick in system power running tighter latencies as Eva suggested and thanks to hipro's excellent strap testing on Striker seen here. Running a little less mhz added around 500-1K to each result and amazingly 18K to aqua, fun stuff.

Have you ever noticed your vcore going up without you changing it your self?
I have the same board and when I set vcore to 1.4v in bios I get 1.37v bios read BUT when I change ONLY my FSB from 400mhz to anything above the vcore changes to 1.47v bios read. I tested all bioses(302, 405, 501, 601) and only with 302 the vcore doesnt change at all.

Have you or anyone else noticed this?

[G H Z]

DI the CPU last night, forgot to cap 03, it was only 73.5K or so.



01


05


06


Aqua

[G H Z]

GHZ how much vdimm you use for 445mhz 3-3-3-8 1T ?

2.5V

445 CAS 3 1T is just nuts.

Not much of a difference between 3-4 from what I can see

Have you or anyone else noticed this?

Never noticed this but I'm not really looking when I'm benching either. If it's stable and completes a run it's good enough for me.