first r420 review leak
- Thread starter Ulukay
- Start date
surfhurleydude
Newcomer
I have this odd feeling that if you're interested in Doom 3, or any games based off of its engine, the X800 series is not for you.
surfhurleydude
Newcomer
Ardrid said:
Well, I've had both a 5950 and my current 9800 Pro. I'd say from my own personal experiences they are pretty much as fast as each other, except one sacrificies IQ for the speed it has.
This time they are pretty much as fast as each other as well, but without the IQ hacking.
As for the Radeon X800 Pro pricing, check the 3DCenter thread.
If we assume 6gpix/s for X800 PRO, then ~50% higher yields 9gpix/s for the X800XT. 9gpix/16 pipes = ~562Mhz =~ 550Mhz. The question becomes, does the X800XT have more than 8 extreme pipes? Do all R420's start out with 16 xtreme pipes, but like with "quad disablement" (16->12) there is also "ALU disablement"? Perhaps in general, many of the extra ALUs are bad, so the challenge is finding atleast 3 quads that are good (12 pipes) and having atleast 16 good ALUs in 2 of the quads (8 extreme pipes), but only 4 good ones in the other?
Does this introduce the possibility of a later X800XT that has 16 full extreme pipelines as yields go up?
Does this introduce the possibility of a later X800XT that has 16 full extreme pipelines as yields go up?
If the Farcry scores are real, Im pretty impressed. Even if the GPU isnt being stressed that much. Almost 20fps more, without the leaked drivers. Farcry has a LOT of issues with NV cards. The 1.1 patch gave them a FPS boost, but caused lots of bugs. Same for the leaked drivers, gave a huge FPS boost, but caused lots of bugs.
Hopefully someone will put up some comparison shots to show the difference in the two. If they even had time to bench it with the leaked drivers.
Hopefully someone will put up some comparison shots to show the difference in the two. If they even had time to bench it with the leaked drivers.
Tom's Hardware is a complete joke. Go back and read his review of the 6800. He benchmarked nine games...all nine of them were "TWIMTBP" games. It's no surprise that Breed and FarCry, "TWIMTBP" games, were used. I guarantee all the games he benchmarked for his X800 review are also "TWIMTBP" games. It is by far the absolute most nVidia biased site on the internet.
DerekBaker
Newcomer
JBark said:surfhurleydude said:
Where did you get that $299 from? Even Hellbinder the crazy ATI mansays $399 for the Pro and $499 for the XT. If I could get an x800Pro for $299, it would defintely sway my new card purchase more towards ATI.
£299.99 ?
Derek
MasterBaiter
Regular
"Meh" is all I have to say. :? 1 more day... I am really interested in seeing the proper numbers though.
trinibwoy said:Ulukay said:
huh? You mean relief? I'd like to see few more higher rez tests before I would make any conclusions.
Also nice to see that THG still hasn't figured out that it's X2 the Threat, not the Thread.
Those scores are really odd. The X800XT has 47% more fillrate and 24% more memory bandwidth, and yet with AA/AF, situations one would expect it to scream past the X800 Pro, it's only barely faster? Does anyone else find that just a bit odd, even at 1280x1024? At 1600x1200 the expected performance delta between the two cards shows up in force, but 1280? Weird..... Who'd have thought 1280 would be cpu limited.
Someone got an Athlon 64?
Someone got an Athlon 64?
Similar threads
- Locked
