anandtech early x1900xtx benchmarks

pandora's box said:
not that big of an improvement considering its compared to a 7800GTX 256MB.

oops, thought it was a 512MB version :(
still - for a 'refresh' I must say it's what I expected.
 
yeah it looks like ati will win the performance crown for now, but i bet the G71 from nvidia will take it back again.
 
ATI has told its partners that the X1900 launch will be an actual product launch; meaning we should see inventory in the stores on the 24th. This may be true for ATI branded inventory, but AIB partners say it will be several weeks before they have reliable shipments of inventory ready for merchants

Oh bollocks. :mad:
 
Also, they used beta drivers from back in the december since the new cats don't recognize the card.
 
wankel said:
BTW...the card tested is the XT not the XTX.


Well the difference between xt and xtx is like 25mhz core 50mhz memory if im not mistaken, talking like 4fps or so maybe. :confused:
 
Nice find! Not sure what it means using beta drivers and such, but nice to see it is working. :)
 
Looks good, anything that makes games faster is a + :D even if its marginal
 
Just wondering, the article mentioned that the X1900 would be using the DVI jumper cable for SLI operation instead of a hard bridge. Does that mean it can be used on any brand motherboard, or does it still require one with the Radeon Xpress chipset? Because I would love two of these on a Asus A8N32-SLI.
 
westrock2000 said:
Just wondering, the article mentioned that the X1900 would be using the DVI jumper cable for SLI operation instead of a hard bridge. Does that mean it can be used on any brand motherboard, or does it still require one with the Radeon Xpress chipset? Because I would love two of these on a Asus A8N32-SLI.

Yes and No, its possible, and internally ATI and a number oems(Voodoo PC's, the owner was talking about how they were running CF in-house on a Asus A8N because the ati presenters forgot to bring a cf motherboard when showing them CF) already ar running it on diffrent platforms such as the nforce, the problem is really drivers on ati's site, if they don't release offical drivers they'll be hacked soon enough.
 
I am kinda new at this but what is the difference between pixel pipelines and pixel shader processors?
 
imo those preliminary numbers are quite dissappointing. hopefully this is only a driver issue and the x1900 should fly once proper drivers are installed
 
Disappointing.

Is a EVGA 7800 overclocked at all out of the box?

Edit: Yeah it's 490 on the core which Anand didn't note. So that makes it almost halfway to a 512 at least, anyway, which makes the X1900 look slightly better.

Edit 2: Actually it's either 450 or 490 seems like.
 
RoffleCopter said:
imo those preliminary numbers are quite dissappointing. hopefully this is only a driver issue and the x1900 should fly once proper drivers are installed

For all those that thought it was dissapointing, what did you expect? The x3 shaders are only gonna to show their strenght in games that make heavy use of shaders. Lots of those scores were run at the higher AA and AF setting which is more fillrate and bandwidth limited then shader limited. The bottle necks of the x1800 and the x1900 are going to swap.
 
yeah, it seems that there's something odd with those benches. But the article doesn't state how they benchmarked. They might have custom timedemos that stress the hell out of the cards. I'll reserve my judgement until launch.
 
is it just me or does it not really bring "much" performance boost ?according to the hype.
 
Notice that the CPU is an Opteron 165 running, from what I can tell, stock at 1.8GHz. I don't know why they would break NDA for that worthless preview.
 
I doubt the NDA was broken. NDAs have terms which can be negotiated on-the-fly, and it seems like ATi PR probably allowed them to publish these results. Giving consumers a little taste before the big roll-out is hardly ever a bad idea, especially when the numbers are actually quite good.

Now, was it a case of ATi stipulating that Anandtech could not release any numbers comparing the XTX/XT to the GTX 512 before the NDA is null and void? I'll say maybe. I have a hard time believing that Kris simply did not have a GTX 512 available when putting these results together.
 
Haha, anyone else get the feeling that ATi was waiting for 3Dmark06 to be released so that they could surprise announce the X1900?
 
ZenOps said:
Haha, anyone else get the feeling that ATi was waiting for 3Dmark06 to be released so that they could surprise announce the X1900?

I think futuremark waited until the x1900 was coming out to release 3dmark06, so there would actually be a card that can run it :)
 
phide said:
I doubt the NDA was broken. NDAs have terms which can be negotiated on-the-fly, and it seems like ATi PR probably allowed them to publish these results. Giving consumers a little taste before the big roll-out is hardly ever a bad idea, especially when the numbers are actually quite good.

Now, was it a case of ATi stipulating that Anandtech could not release any numbers comparing the XTX/XT to the GTX 512 before the NDA is null and void? I'll say maybe. I have a hard time believing that Kris simply did not have a GTX 512 available when putting these results together.

This is DailyTech, not Anandtech. They're different companies now. Kris said in the forum he doesn't sign NDAs and his results have nothing to do with Anandtech. The only reason he got a card ahead of the launch is because he got someone in Taiwan to express ship him one.

Anandtech will have the review up on Tuesday, when the NDA expires. Any benchmarking done before then is by people who haven't signed an NDA but somehow got a card.
 
Ah. Thanks for the correction. I suppose I was under the impression that DailyTech was still technically under the AnandTech umbrella (and they'd thusly have to adhere to AnandTech NDAs).
 
Back
Top