6800GT preview up at guru3d, link inside

[James3shin]

Link

 

[Matthias99]

link to first page of article instead of the third

 

[James3shin]

tee hehe sorry i linked where i left off, btw credit for this find goes to the guys at nvnews.net and the guys at guru3d for writing the article

 

[GeneralGrievous]

The GeForce 6800 GT at default clock speeds is doing 350 MHz core and 2x500 MHz memory (1000 effective).
overclocked it was capable of running of a 401 MHz core and 530 MHz (1060 effective) memory frequency.

Not exactly a great overclocker. Looks like this might be an actual retail GT as opposed to a handpicked Ultra core slapped on a GT for reviewers.

 

[James3shin]

even without OCing it franky held its own quite convincingly.

 

[DAPUNISHER]

Originally posted by: James3shin
even without OCing it franky held its own quite convincingly.

Damned straight. Besides If I can get ultra level performance from my GT while having a single molex and slot cooler I won't be complaining. Especially if the $317.99 to my door gets honored :beer:

 

[Rent]

Originally posted by: GeneralGrievous

The GeForce 6800 GT at default clock speeds is doing 350 MHz core and 2x500 MHz memory (1000 effective).
overclocked it was capable of running of a 401 MHz core and 530 MHz (1060 effective) memory frequency.

Not exactly a great overclocker. Looks like this might be an actual retail GT as opposed to a handpicked Ultra core slapped on a GT for reviewers.

Thats one of those YMMV type things. I'm sure some people will get GTs that overclock like hell.

 

[GeneralGrievous]

I'd like to see some comparisons to a solidly overclocked x800 Pro.

 

[Killrose]

I would just like to se an actual 6800GT on the store shelves

 

[nemesismk2]

Originally posted by: GeneralGrievous
I'd like to see some comparisons to a solidly overclocked x800 Pro.

It would have to be to beat a 6800GT!

 

[CrazyHelloDeli]

Originally posted by: Killrose
I would just like to se an actual 6800GT on the store shelves

Exactly. These paper launches are really hurting the industry IMO.

 

[CaiNaM]

"As Antialiasing results where not consistent between different brands we can't use them in the part of the benchmark results."

wth does THAT mean???

 

[kylebisme]

well they are trying to say that since ati and nvidia have different aa, they should ignore benchmarking any aa at all. but what it means is that they are morons and we should simply disregard the whole review.

 

[vapidtransit]

I'm seriously thinking of cancelling my x800XT order and going after one of these.

 

[Genx87]

GT is smoking that X800 Pro. Hell the thing is giving the X800 XT a run for its money in some of the benches.

 

[Ackmed]

Am I missing something, or did they not test with AA?

They really need to take up more space on the page in their reviews. I dont like the tiny column like section down the middle, with adds flashing in your face on the sides.

edit, I didnt see that part, cainam. I too dont understand their problem with using AA.

 

[GeneralGrievous]

The companies antialiasing performance seems to be about the same, so I guess they can cut it out from both cards. Turn it on and both have similar percentage drops.

Hell the thing is giving the X800 XT a run for its money in some of the benches.

And the x800 Pro gives the 6800 U a run for its money in some benches. Of more popular games I might add.

 

[Ackmed]

This review has some wierd anomolies in it. They dont have AF scores for the X800 Pro in Splinter Cell. And they claim the AA "problems" are a known "issue". Its not a issue, the game doesnt support AA.

Im not sure how anyone can cliam the GT is "smoking" the Pro. Most benches are within a few frames in 1280x1024 with some AA/AF. UT2004 is in the teens faster on the GT than the Pro though.

Another review that makes no mention of the bugs in Farcry with NV hardware.

 

[GeneralGrievous]

UT2004 is in the teens faster on the GT than the Pro though.

Which is the exact opposite of the firingsqaud review; there the Pro was leading the UT2k4 benchmark by a considerable margin.

 

[Ackmed]

I would just like to see, A) GT's on the street for sale, B) Final drivers and retail boards reviewed(MSI, eVGA, BFG, etc.) C) Reviews from reputable sources, like HardOCP, Anand, B3D, etc.

 

[Zebo]

I hate review site you can't print. Also yet another review site testing on a slow p4 system instead of gamers and NVidia choice the A64. Blah.

It does look like this is *the* card to get though.

 

[Zebo]

Originally posted by: Ackmed
Am I missing something, or did they not test with AA?

They really need to take up more space on the page in their reviews. I dont like the tiny column like section down the middle, with adds flashing in your face on the sides.

edit, I didnt see that part, cainam. I too dont understand their problem with using AA.

I did'nt have any ads or flashing. Download a good HOSTS file and AVANTso you don't have to put up with these people peddling thier warez and wasting your bandwitdh.

 

[Keysplayr]

Looks like it is inevitable.

 

[CaiNaM]

Originally posted by: Genx87
GT is smoking that X800 Pro. Hell the thing is giving the X800 XT a run for its money in some of the benches.

:roll:

 

[CaiNaM]

Originally posted by: Ackmed
This review has some wierd anomolies in it. They dont have AF scores for the X800 Pro in Splinter Cell. And they claim the AA "problems" are a known "issue". Its not a issue, the game doesnt support AA.

Im not sure how anyone can cliam the GT is "smoking" the Pro. Most benches are within a few frames in 1280x1024 with some AA/AF. UT2004 is in the teens faster on the GT than the Pro though.

Another review that makes no mention of the bugs in Farcry with NV hardware.

fanboys will be fanboys.... :disgust:

where in this review did the gt "smoke" anything? :shocked:

i've only seen 1 title where, with the lastest fw drivers, the gt has a pretty commanding lead: CoD (and it wasn't included in this review), but even then the slower PRO runs it at 55fps @ 2048x1536 4xaa/8xaf... hardly what one would consider anywhere near 'slow'.

but rather than argue further, i'll let the fanboys have their fun. it's a safe bet that these cards will go back and forth regarding performance with future drivers. they're close enough that the differences are minimal, and there is plenty of room for driver improvements on both sides, even the "old" r420 architecture - during the "trilinear" optimization PR chat (and it's PR as imo ati was clearly attempting to "get away with one" in this matter) one of the ati engineers stated the new memory architecture of r420 was only running around 70% efficiency.

the only real trump card nvidia has is sm3, which may or may not prove to be of any benefit in this generation of cards.