PDA

View Full Version : 6800GT review at trusted reviews


Pages : [1] 2

Vapor Trail
06-21-04, 01:22 PM
The 6800 Gt is looking better all the time!!! :)


http://www.trustedreviews.com/article.aspx?art=503

3madcows
06-21-04, 02:59 PM
Regarding Far Cry's AA problem on the X800pro

"However it’s telling that at 1,024 x 768 in 4x FSAA and 4x AF the 6800 GT is still faster than the X800 Pro, even though the latter card was clearly not running with these effects enabled."

Man, that's awesome.

Sazar
06-21-04, 03:08 PM
comparing a 16 pipe card to a 12 pipe card don't you think those results are kinda expected?

:confused:

Vapor Trail
06-21-04, 03:11 PM
comparing a 16 pipe card to a 12 pipe card don't you think those results are kinda expected?

:confused:


Yes, but early previews showed the exact opposite.

Looks like Nvidia's drivers are starting to come around and I suspect the 1.2 patch will help even more. :)

Sazar
06-21-04, 03:16 PM
the initial reviews showed the x800pro being slower than the 6800ulra by a little to a larger amount... keeping neck and neck in some situations but slower in most others...

it is expected therefore that a GT which is in effect just a slower ultra should be able to be faster in general than the x800pro :)

heck the x800pro being able to keep up with fewer pipes is a decent indication of its performance... but @ the same price point its clear which would be taken...

-edit-

its weird they tested far cry on the different cards using different patches :confused:

why 1.0 for one and 1.1 for the other?

Vagrant Zero
06-21-04, 03:16 PM
comparing a 16 pipe card to a 12 pipe card don't you think those results are kinda expected?

:confused:

Not to Atidiots, thanks to [H]OCP some of the more vocal [read annoying] ones were proclaiming that the X800Pro was faster than the ULTRA!

Sazar
06-21-04, 03:22 PM
Not to Atidiots, thanks to [H]OCP some of the more vocal [read annoying] ones were proclaiming that the X800Pro was faster than the ULTRA!

there is no such thing as an atidiot firstly :)

secondly there are tests where the card indeed was coming out faster than the ultra... but given the whole picture its a little hard to imagine a 12 pipe card beating a 16 pipe card with the raw power of the ultra... as we have seen from members on these forums itself who actually HAVE the high end cards there is no much between the top 2... the pro is slower but it still has the juice to handle the games :)

as more people get their hands on the GT on this board it'll be nice to see them post their findings... I am personally waiting for saturn to post his results so we can compare em with jakUP and riptides and arioch and company...

particleman
06-21-04, 03:22 PM
I do think the 6800 GT is a better buy than the X800 Pro (I've never been a big fan of pipeline crippling), but personally I that reviewer is doing something wrong, as I've seen plenty of screenshots posted around on forums and reviews that showed AA and AF do work on the X800 Pro in FarCry.

Sazar
06-21-04, 03:26 PM
he apparently tested using different patches for the 2.. there is no indication he used patch 1.1 for the ati card while using it for the nvidia card... this should theoretically make little difference but he was saying he saw similar anomalies on both platforms and the nvidia one cleared up after the patch...

I personally saw no difference between going from one patch to the next (seeing as how the r420 is derived in part from the r300 core) and I know that AA seems to be working fine given the results of everyone else I know to have tested the product...

it gets curiouser and curiouser :cool:

/me got that line from some strange B flick he can't remember the title of...

Vagrant Zero
06-21-04, 03:30 PM
Carroll Lewis's Alice's Adventures in Wonderland (1865) is where the line first appeared.

And one or two benchs doesn't make a card faster. The Pro doesn't compete with the Ultra [that is after all what the GT is there for], especially after all these new driver releases.

anzak
06-21-04, 03:31 PM
comparing a 16 pipe card to a 12 pipe card don't you think those results are kinda expected?

:confused:

No, The GT may have more pipes but the pro has a much higher clock speed. Just look at the fill rate:

6800 GT 350*16 = 5600 mega pixels per second

X800 pro 475*12 = 5700 mega pixels per second

Gogeta
06-21-04, 04:20 PM
No, The GT may have more pipes but the pro has a much higher clock speed. Just look at the fill rate:

6800 GT 350*16 = 5600 mega pixels per second

X800 pro 475*12 = 5700 mega pixels per second

thats what Ive been saying. also looking at the 6800 NU

325x12 = 3900 mp/s, about the same as a slightly overclocked 9800 XT....and people still say you cant compare 9800 to 6800nu, you we can.

anzak
06-21-04, 04:26 PM
thats what Ive been saying. also looking at the 6800 NU

325x12 = 3900 mp/s, about the same as a slightly overclocked 9800 XT....and people still say you cant compare 9800 to 6800nu, you we can.

Your right and err... wrong at the same time. 3900/8 = 487mhz! Thats not a small overclock.

Ruined
06-21-04, 05:57 PM
comparing a 16 pipe card to a 12 pipe card don't you think those results are kinda expected?

:confused:

Considering both the 16-pipe 6800GT and 12-pipe X800PRO have exactly the same MSRP ($399) its a very fair comparison.

c4c
06-21-04, 06:01 PM
No, The GT may have more pipes but the pro has a much higher clock speed. Just look at the fill rate:

6800 GT 350*16 = 5600 mega pixels per second

X800 pro 475*12 = 5700 mega pixels per second

Thank you for bringing some sanity to this thread.

More pipes does not automatically mean more fillrate :)

JBark
06-21-04, 07:30 PM
Thank you for bringing some sanity to this thread.

More pipes does not automatically mean more fillrate :)

Exactly, but more pipes do mean a larger perfomance increase when overclocking. Overclock each of them 25MHz, and now they are equal:
6800GT 375*16 = 6000Mp/s
x800Pro 500*12 = 6000Mp/s

Increase that by 25 again, and now the 6800GT has the lead:
6400Mp/s vs. 6300Mp/s.
Not a huge gain with 16 vs 12 pipes, but it is a 200Mp/s swing in the 6800GTs favor.

Clay
06-21-04, 08:54 PM
Ah, the wonders of multiplication. :angel: Just kidding, what you say is true of course...just really obvious (well, it should be at least). Having said that probably a good thing that you spelled it out. :)

Nv40
06-21-04, 08:56 PM
comparing a 16 pipe card to a 12 pipe card don't you think those results are kinda expected?

:confused:

not for the people that still link to (cough) [H] reviews.. or others style of reviews ..like [DH].. which they recommends the X800 and the XT againts any NVIdia card .. :rolleyes: they can't use 60.72 drivers forever ... and allowing ATI to use all their "optimizations".. which there are many (cough) "they way gamers will play".... or using lower setting here and higher there... a tell gamers which one is "the fastest" or the better card for them.. :screwy:

Gogeta
06-21-04, 08:58 PM
Your right and err... wrong at the same time. 3900/8 = 487mhz! Thats not a small overclock.

ok, wise guy. Say 450Mhz, its a close match nonetheless.

c4c
06-21-04, 11:21 PM
Exactly, but more pipes do mean a larger perfomance increase when overclocking.
Naturally :D

Overclock each of them 25MHz, and now they are equal:
6800GT 375*16 = 6000Mp/s
x800Pro 500*12 = 6000Mp/s

Increase that by 25 again, and now the 6800GT has the lead:
6400Mp/s vs. 6300Mp/s.
Not a huge gain with 16 vs 12 pipes, but it is a 200Mp/s swing in the 6800GTs favor.
I think you meant a 100Mpix swing in the GT's favor...What is that, something like a 1.5% difference in fillrate? Somehow I dont think that will make or break the benchmarks. Regardless, even with a (small) fillrate deficit at stock speed the GT is shaping up nicely.

Of course, this assumes everybody will get similar overclocks...Or they will choose to overclock at all. Personally I'd rather stick with guaranteed numbers. I mean a 50 Mhz overclock on a GT is still only like a 14% gain..

dan2097
06-22-04, 05:10 AM
The 6800GT has 11% more memory bandwidth which is probably why it generally holds the lead.

Fill rate is~equal

The 6800GTs more powerful pixel shaders and the X800 pros faster vertex shaders probably ~cancel out leaving you with just the 6800GTs memory bandwidth advantage.

Of course thats a gross simplification.

jbirney
06-22-04, 09:30 AM
Well I am glad Sazar can think on his/her feet. Why in a shader intestive game would you think just fill rate is the ONLY limiting factor here. The GT is fast as it has more total shader units and on shader limited game.... well fill rate is not the be all end all. Maybe lets use some common sense here? Plus the GT has more MB so its a DUH it should win all the time type thing....

TheLostSwede
06-22-04, 09:31 AM
Just to clear things up here, the X800 Pro pre production card was indeed tested with an unpatched version of Far Cry as the test was not done at the same time as that of the 6800 GT. We do however have a X800 Pro retail card in now and should have the review up of that in a weeks time or so

/Lars

DivotMaker
06-22-04, 09:41 AM
not for the people that still link to (cough) [H] reviews.. or others style of reviews ..like [DH].. which they recommends the X800 and the XT againts any NVIdia card .. :rolleyes: they can't use 60.72 drivers forever ... and allowing ATI to use all their "optimizations".. which there are many (cough) "they way gamers will play".... or using lower setting here and higher there... a tell gamers which one is "the fastest" or the better card for them.. :screwy:

:rolleyes:

Lfctony
06-22-04, 10:18 AM
The 2 cards are priced the same, so does make sense to compare them directly.