Not thinking in terms of victory or defeat, the ATI's new architecture seems very forward looking. First, alot of people suspected this was just R420 on steroids which was R300 on steroids, but that was not true. Second, the g70 is basically the nv30 architecture which was a failure at first and then matured and now is impressing everybody. So the fact ATI has managed to be more than competitive in certain games while dealing with new architecture is impressive. THis is assuming they aren't cheating though like the 5 series NV's were. ANd of course 3dmark isn't a game but ATI'S STRONG showing in that along with other newer games like FEAR might portend a bright future for r520 based tech. So while if you are buying this generation you should probably go with Nvidia based on price alone, it is very possible by the time r580 and better drivers come around ATI might be pulling strongly ahead of Nvidia. ON the other hand, WTF is with ati and open gl? They've had so many years to work with it and still can't get it right?
I have to agree. I like a lot of the ideas presented here. ATi, as AMD has shown us, is looking in the right direction by aiming to create a more efficient architecture based around things like a thread dispatch unit and ring bus. They're both good ideas with a lot of promise, and I'm glad ATi is going in that direction.
What I'm not happy about is the delayed launch, the lack of availability, the nonexistent crossfire, and all the "Just a little longer!" PR releases. Either deliver your product or shut up.
I like the architecture, but ATi has some serious patching up to do with the public.
I was just wondering about how many of you guys have read any other website thats postted benchmarks.
To me the general concensus from the other sites is that X1800 does perform much better than the GTX in higher res games and with AA and AF enabled.
When AA and AF are not enabled or either one is not enabled the performance lead is cut to the GTXs level or lower. I was astounded by the benchmarks on this site and how lacklustre the review was and the writing of it. All it pretty much was was loads of info on the architecture and nothing on the benchmarking side, and then a conclusion.
When benchmarking it would be nice to be told the AA and AF numbers. The high end stuff was done in 4xAA... Now come on it should be able to hand 16xAF and 4xAA. And look at the whopping number of games benched. Other sites have up to 10 games and sometimes more to bench. We get 5 where 1 of them was used on only 1 card for the really high res stuff. Great. This review had TONS of info that i really wanted to see.... now ill be off to read some other benchmarks from other sites that actually have done some work.
To me the general concensus from the other sites is that X1800 does perform much better than the GTX in higher res games and with AA and AF enabled.
When AA and AF are not enabled or either one is not enabled the performance lead is cut to the GTXs level or lower. I was astounded by the benchmarks on this site and how lacklustre the review was and the writing of it. All it pretty much was was loads of info on the architecture and nothing on the benchmarking side, and then a conclusion.
When benchmarking it would be nice to be told the AA and AF numbers. The high end stuff was done in 4xAA... Now come on it should be able to hand 16xAF and 4xAA. And look at the whopping number of games benched. Other sites have up to 10 games and sometimes more to bench. We get 5 where 1 of them was used on only 1 card for the really high res stuff. Great. This review had TONS of info that i really wanted to see.... now ill be off to read some other benchmarks from other sites that actually have done some work.
i think its an incredible card for 16 pipes HOT *sizzle*. Now i get to live out my dream of being on Best Week Ever!!!!!
x1800 - HOT
x1600 - not
x1300 - not
From a technological perspective, the x1800 is way ahead of the gtx. It runs SM3 like Nvidia only wishes it's cards could run it. And it has better IQ. The only problem is you cant buy one. OpenGL performance is also lacking, but that can be fixed in the drivers. I have a feeling in Unreal3 it's gonna pwn the 7800.
x1600 - not
x1300 - not
From a technological perspective, the x1800 is way ahead of the gtx. It runs SM3 like Nvidia only wishes it's cards could run it. And it has better IQ. The only problem is you cant buy one. OpenGL performance is also lacking, but that can be fixed in the drivers. I have a feeling in Unreal3 it's gonna pwn the 7800.
I think in UT2007, both the 7800GTX and the X1800XT will both get pwned by the game's advanced graphical engine. I would think you would need R580 or G72/G80 (whatever it's called) to play that at high res and with AA/AF.
Even F.E.A.R., which looks to be much less advanced than the Unreal 3 engine, gives these two graphics cards a real workout . . . and F.E.A.R. is a game that will be released BEFORE the X1800XT . . .
I've read that X1800XL overclocks to XT's speeds easily, it will be interesting to see the oc results on that thing.
Yeah, that's what I would think since it's supposed to be just a throttled version of the XT. But I'm wondering if they're using the same RAM (Samsung 1.26ns) since the XL is running it's ram 500 mhz slower . . .
Most of the stuff I read (HardOCP, Hexus, Tech Report) was in line with ATs findings....with the requisite few differences, of course.
X1800XL is not HOT considering it clearly loses to a 7800GT. It may be warm, but not HOT. X1800XT will be the only HOT thing there...
I'm still reserving judgement. I need to see:
-if there will be a myriad of cards filling the gap between 1800XT and the 1600 series, based on the 520 core and 256 bit mem. Similar in fashion to the X800GTO^2
-how will the XL overclock once the supposed cores that didn't make the speed grade are liquidated
-what the market price for the XL or card just below the XL (if one is ever released based on the 520 core) winds up being.
Pluses
-like the focus on IQ and features. somewhat of a role reversal between the rehash 420 and the 6800 series from a feature aspect.
-like the high clocks mainly because I just like high clocks. If more pipes are utilized in a near future release, at least there is high probability that the clocks will be there.
-like the movement to 90nm and ring bus. 90nm seems to have it's problems across the board for everyone who moves there. That hurdle is crossed. Don't know much about chip architecture, but any innovation is a good thing.
Minuses
-1600 series is not highly competative and I usually shop the two-step below flagship card models. Hope there will be some in between models using the r520/256bit pcb.
-overhyped and late. Not much desire to splurge and buy any of these.
-unavailbility, a minus for some if you were incline to get one.
-performance is not overwheming for such a late release.
Will I dump my X800Xt today on either this or a 7800 series-------------no, not just yet.
-if there will be a myriad of cards filling the gap between 1800XT and the 1600 series, based on the 520 core and 256 bit mem. Similar in fashion to the X800GTO^2
-how will the XL overclock once the supposed cores that didn't make the speed grade are liquidated
-what the market price for the XL or card just below the XL (if one is ever released based on the 520 core) winds up being.
Pluses
-like the focus on IQ and features. somewhat of a role reversal between the rehash 420 and the 6800 series from a feature aspect.
-like the high clocks mainly because I just like high clocks. If more pipes are utilized in a near future release, at least there is high probability that the clocks will be there.
-like the movement to 90nm and ring bus. 90nm seems to have it's problems across the board for everyone who moves there. That hurdle is crossed. Don't know much about chip architecture, but any innovation is a good thing.
Minuses
-1600 series is not highly competative and I usually shop the two-step below flagship card models. Hope there will be some in between models using the r520/256bit pcb.
-overhyped and late. Not much desire to splurge and buy any of these.
-unavailbility, a minus for some if you were incline to get one.
-performance is not overwheming for such a late release.
Will I dump my X800Xt today on either this or a 7800 series-------------no, not just yet.
mooncancook
Platinum Member
- May 28, 2003
- 2,874
- 50
- 91
The X1800 is just warm because of lackluster performance in OGL. 7800 is not hot either because of all those IQ issues (shimmering and blocky shadow texture). But I think the architecture of X1800 is HOT! Just imagine what it can do with 24 or 32 pipes if ATI can ever fab it.
That was only becuase of all the ATi fanboys that flooded in the last few weeks bombarding us with Inquirer articles about rumored 3dmarl scores, core speeds, etc.....It was all hype...
flexy
Diamond Member
- Sep 28, 2001
- 8,464
- 155
- 106
expected outburts of enthusiasm - but got a cold chill of reality.
The wonder card arrived - but too bad years too late. And too prizey. And of course its not available yet anyway.
Life goes on. [pets my X850XT
]
Wake me when we have R580 or similiar
The wonder card arrived - but too bad years too late. And too prizey. And of course its not available yet anyway.
Life goes on. [pets my X850XT
]Wake me when we have R580 or similiar
It doesnt matter if the results are not there.
The NV30 was technologically superior too, and it lost as well.
I'm glad people are recognizing the OGL issues, they promised a long time ago to fix that.. as you say "in the drivers".
Your excuse it can be fixed in the drivers is not going to cut it. They've had years.
Certainly not the first batch! So,if you wait till January after the old stock clear out in Dec shopping season, the xl could be a good OC part.
I think it may depend largely on overclocking ability.
The r520 allows voltage modification via software, a huge plus. If it actually has better image quality, and can get a higher boost from overclocking than the GTX, I might consider one.
Otherwise, if the rumors are true about nvidia holding back perf. until the 80.xx drivers, plus dual core optimizations, I'm sticking with the GTX. Sure, R520 might have dual core optimizations ... in 6 months
The r520 allows voltage modification via software, a huge plus. If it actually has better image quality, and can get a higher boost from overclocking than the GTX, I might consider one.
Otherwise, if the rumors are true about nvidia holding back perf. until the 80.xx drivers, plus dual core optimizations, I'm sticking with the GTX. Sure, R520 might have dual core optimizations ... in 6 months
apoppin
Lifer
x1800xt? . . . as "hot" as the 7800GTX . . .
just different features . . .
very very competitive
what i expected
No AGP version . . . sad for me or i would definitely buy one.
just different features . . .
very very competitive
what i expected
No AGP version . . . sad for me or i would definitely buy one.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter