8800GT VS. 8800 Ultra, scores out!

[trans am]

is this real? o/c the core 300mhz from 600mhz stock?

[mascaras]

Default core clock = 600mhz

the 900Mhz its already "Overclok" (GPU-Z SS)

[RPGWiZaRD]

3dmark 2005 ??????

14K in 2005 its "nothing" , maybe they meant to say 3dmark 2006 or maybe they run 3dmark 2005 in another resolution than the default


INQ...........................


btw:



cumps

LOL, INQ is at it again with their typos and what not. Must be the that they are too hooked trying to see what the magic crystal ball displays that they forget what they're writing about.

So 1.5GHz and 112 SPs confirmed now by NVIDIA, phew.

[trans am]

default = 600mhz

"Overclok" (GPU-Z SS = 900mhz)

lol Yeah I just noticed that and revised the post. Any idea what cooling was used?

If thats just water or air 300mhz is pretty amazing!

also you can see in GPU-z the stock 1500mhz shader. very nice!

[mascaras]

well it would be nice if they used stok cooling

yep , nvidia already confirmed the 1500mhz shader clock

[trans am]

well it would be nice if they used stok cooling

considering people popping champagne from getting 700 mhz gpu from a 8800gts stock of 500mhz core clock.

Yeah!

[ripken204]

we need benchmarks with AA OFF and also some OCed results. my 8800GTS at stock is beating the 8800GT's results and when its OCed it is the same as the ultra..

[DMH]

we need benchmarks with AA OFF and also some OCed results. my 8800GTS at stock is beating the 8800GT's results and when its OCed it is the same as the ultra..

Remember the result's of GT being with a X6800 and yours with a Q6600(2-core vs 4-core)

[trans am]

we need benchmarks with AA OFF and also some OCed results. my 8800GTS at stock is beating the 8800GT's results and when its OCed it is the same as the ultra..

is that with the opteron or the quad?

[TreeTop]

we need benchmarks with AA OFF and also some OCed results. my 8800GTS at stock is beating the 8800GT's results and when its OCed it is the same as the ultra..

Sorry but the $400 plus you dropped on the GTS is about to get ownd by a $250 card

[ripken204]

i have a Q6600 but i'm not counting in the cpu score, i find it stupid that you guys always do that. i'm comparing the 2 scores for the graphics card. and i have the 8800GTS 320MB, the cheapest EVGA one, which i got for $280 off of newegg.

edit: i just re-ran 3dmark06 and you guys can see that my video card while OCed beats the ultra at stock. and my Q6600 is only at 3ghz right now and theirs IS with a Q6600 at 3.6ghz

[TRF-Inferno]

i have a Q6600 but i'm not counting in the cpu score, i find it stupid that you guys always do that. i'm comparing the 2 scores for the graphics card. and i have the 8800GTS 320MB, the cheapest EVGA one, which i got for $280 off of newegg.

edit: i just re-ran 3dmark06 and you guys can see that my video card while OCed beats the ultra at stock. and my Q6600 is only at 3ghz right now and theirs IS with a Q6600 at 3.6ghz

Run it again with 4xAA and 8xAF, just like what that reviewer did. You also need to realize that while your card may give an Ultra a run for its money, that Ultra is still running stock clock speed. They simply plug it in, install drivers and that's it. You had to spend some time overclocking your GTS to a safe level, and probably used Rivatuner or ATiTools to apply those clocks when Windows starts up.

[Jodiuh]

i have a Q6600 but i'm not counting in the cpu score, i find it stupid that you guys always do that...8800GTS 320MB...my video card while OCed beats the ultra at stock...

I find it stupid that folks use a benchmark that spits out a number to compare game performance. Play World in Conflict much? It's not all about the number.

[DilTech]

i have a Q6600 but i'm not counting in the cpu score, i find it stupid that you guys always do that. i'm comparing the 2 scores for the graphics card. and i have the 8800GTS 320MB, the cheapest EVGA one, which i got for $280 off of newegg.

edit: i just re-ran 3dmark06 and you guys can see that my video card while OCed beats the ultra at stock. and my Q6600 is only at 3ghz right now and theirs IS with a Q6600 at 3.6ghz

They ran it with AA/AF...

[mascaras]

Remember the result's of GT being with a X6800 and yours with a Q6600(2-core vs 4-core)

Negative , he´s talking about the 8800GT score with Q6600 @ 3600mhz but they used AA x4 & AF

NVIDIA GeForce 8800GT Vs. 8800 Ultra, 3DMark06 Comparison

The guys over at Expertview are sending new set of benchmark scores, this time comparing the GeForce 8800 GT video card to its big brother GeForce 8800 Ultra 768MB. Using Intel Core 2 Quad E6600@3600MHz, nForce 650i Ultra motherboard, 2GB of DDR2 memory at 800MHz with CL5-5-5-15 timings, ForceWare 167.26 beta and Windows XP SP2, the 512MB version of GeForce 8800 GT have reached 9648 marks while the GeForce 8800 Ultra 768MB video card have topped out 11611 marks. Both 3DMark06 runs are performed using 1280x1024 resolution with 4x AA and 8x AF. The GeForce 8800 GT is set to be released on October 29th.





http://www.techpowerup.com/?41921

[astrallite]

That SM3.0 score is awfully low for the 8800GT, looks a bit weird.

With a lower bandwidth than the 8800GTS but with a ton more shader power (pretty much GTX level) I would've expected roughly identical SM 3.0/SM2.0 speeds (which is basically where the GTX is at).

[astrallite]

Different drivers, different CPUs, but assuming performance scales equally (which I doubt, but this is as close to a direct comparison I can come to):

3dMark06 @ 1280x1024 4xAA/8xAF

8800 GT (512MB) and 8800 GTS (320MB) vs 8800GTX

All scores are ratios of stock 8800GTX performance.

SM2.0 - 8800GT - 85.8%
- 8800GTS - 72.4%
- EVGA 8800GTS OC - 80%
- XFX 8800GTS Fatal1ty - 86%

SM3.0 - 8800GT - 76.1%
- 8800GTS - 72.7% (stock)
- EVGA 8800GTS OC - 80%
- XFX 8800GTS Fatal1ty - 92%

Of course, since these were not done on the same test systems, CPU does take some effect as well as driver differences.

Either way the 8800GT (on a single isolated benchmark of course) is performing as well as a $340 OCed card at a $239-249 price point. And with 512mb of RAM it's going to be a faster card at 2048x1536 and above for most games...assuming either card is even playable at that res :P

[BulldogPO]

Does anybody know what waterblock will fit on these?

[perkam]

Wait for:



Perkam

[Jodiuh]

Well, it looks like the etailers are prepping to unload this weekend.
8800GTS 320MB for $239AR

EDIT: even better... $225 from the egg

[RAW-Raptor22]

considering people popping champagne from getting 700 mhz gpu from a 8800gts stock of 500mhz core clock.

Yeah!

B@stards.... mines only stable to 692...

[astrallite]

is this real? o/c the core 300mhz from 600mhz stock?

I think that's just a GPU-Z bug not recognizing the clock rates under new drivers. With the 167.26 drivers on my Fatal1ty it says under stock core and shader 650mhz and 1600mhz.

Then under core it says 1000mhz and 2462mhz shader, even though I never overclocked this card before. Rivatuner also cannot recognize the clock rates under the new beta driver.



I'm pretty certain under stock cooling both GT and GTS would completely fry at 2.3GHz shader clock =P

[RAW-Raptor22]

I think that's just a GPU-Z bug not recognizing the clock rates under new drivers. With the 167.26 drivers on my Fatal1ty it says under stock core and shader 650mhz and 1600mhz.

Then under core it says 1000mhz and 2462mhz shader, even though I never overclocked this card before. Rivatuner also cannot recognize the clock rates under the new beta driver.



I'm pretty certain under stock cooling both GT and GTS would completely fry at 2.3GHz shader clock =P

Either holy cow McMother of god...

Or:

Paste-up...

[astrallite]

Lol, the funny thing is, if you took the new 8800GTS (supposed 128SP @ 1500mhz) and jacked it up to 2400MHz (as benchzowner was originally stating about a high end this november @2400mhz shader clock), you'd pretty much have a 1 TFLOP card

Of course I assume to actually pull that off you'd need some watercooling or a triple-slot fan solution, lol.

[kite7]

Actually I doubt the card is capable of reaching that kind of speed anyways so it doesn't matter what kind of cooling you put on it