8800GTS G92 quick test on 3dmark06

[Chri$ch]

My XFX 8800GTS arrived today, short test with stock cooler...



my card has 1.20v 3D voltage (measured with DMM)

[Scimitar]

Impressive performance for only using a dual core cpu.

[T_M]

1.31V idle - 1.397V highest load in GT2
Stock cooler

[Chri$ch]

stock voltage (1.20v 3D), card watercooled with EK Waterblock. Diver @ default...

[SKYMTL]

Can you bench that with a game? I would love to see how that translates into REAL WORLD performance please. Maybe Crysis or World in Conflict...

[T_M]

Cooling certainly made a difference there Crisch

[mascaras]

My XFX 8800GTS arrived today, short test with stock cooler...



my card has 1.20v 3D voltage (measured with DMM)

Not bad!!

almost same score as 8800GT but with Quad-core @ 3780mhz , however with 8800GT only 700/950Mhz



8800GT (700/950) + Q6600 @ 3780mhz

[T_M]

1.36V idle - 1.458V max load - Stock Air
(note 860 core is same as 870 core)

[Gris]

18,293! - you gotta be happy with that!

[T_M]

You'd think so, but no
Expect to see much more

[Eddie3dfx]

TM thats amazing, 18k on a single card.
What is holding the core/shader back.. is it the temps or voltage?
What kind of temps are you getting?

[over@locker886]

I got 13200 in 06 at default clocks.(670/1675/972) ..E6700@3600..

vgpu : idle 1,168v , vmem: 1,98v

[Chri$ch]

vGPU ~1.27v idle / ~1.32-1.33v load

[Chri$ch]

Btw, every who want to use Rivatuner with 8800GTS G92 must add "G92 = 600h" in the RivaTuner.cfg

[T_M]

Great work thanks Crisch!
And nice water clocks there for the volts.

At 1.36V idle on stock air 100%, i get 38C idle in 24C room ambients.

[guess2098]

here is some funny test results

[tranceaddict]

Those results make no sense at all, I call it BS!!

[Achill3uS]

nice work Denny

and nice find Crisch, thanks for share

ive picked up today another card, and it could easy 840mhz for 3dmark at stock by the way i can tell now, its BFG OC edition.

[Jodiuh]

It's like NV's drivers activate/deactivate the sp's randomly. NDA's up today or tomrrow, right?

[sonofander]

so does this card outperform the GTX?

[Eddie3dfx]

http://www.xtremesystems.org/forums/...d.php?t=169059

Performance is rather pathetic for the price actually

[guess2098]

Those results make no sense at all, I call it BS!!

how about you go get a card and test it, show results to me

http://www.xtremesystems.org/forums/...d.php?t=169059

Performance is rather pathetic for the price actually

and plus they need to fix driver when AA enable soon
i think my driver (from evga package CD) is too old ....

[T_M]

Guys, this GPU runs REALLY cool!

Stock heatsink at 1.52-1.54V load in 3DMark'06 GT1 + GT2:

[Eddie3dfx]

I don't mean to knock the gts cards at all, I just am saying the difference is so tiny in the games.
I think this is the most fair way to go about it.
Download the newest drivers in xp (169.21)
How about you lower your cpu to 3344, which is the equivalent to my quad at 3520, since yours is a penryn and 5% faster at the same speed as mine.

Then overclock your 8800gts to the highest it can go, I will do the same with my gt and then lets run the same game benchmarks.
This way we are benching the max on both cards with generally the same cpu core and see the difference in the two. Don't get me wrong, If I felt it was worth the difference, I would send my card in to upgrade tomorrow.\

So far the benchmarks I have seen outside of 06-3dmark, shows very little difference between the gt and gts... But most of these have their cores/shader overclocked at small amounts.
Now these benchmarks aren't showing xtremeforums the comparison between a max gts and gt. Thats really what I want to see.
Watercooled


Here is a crysis gpu benchmark thread that I started
http://www.xtremesystems.org/forums/...d.php?t=165411

[Scimitar]

Here is an interesting result I found in the Hardware Canucks review.