DOM:confused: the powercolor is at full 640 sp ?

Yes, the PowerColor card is the proper card there.

DrPepperand lets not forget the die shrink :p

Not all 9800 GT cards use G92b. The partners choose not to mention it at all. Same with 55nm 9600 GT.

nvidia makes some good cards -- there no was no reason they had to do this. you don't need to lie to sell your product, let the benchmarks speak for themselves

Lets face it, Nvidia has been doing this since ATI stepped on the scene, it works well to make the fanboys gloat over minor things and gives them other things to talk about than the actual facts and hard data. It is a excellent business practice against a smaller competitor who has less funds to advertise and counter this BS. It also makes noobs feel that Nvidia is the way to go as they seem the more appealing company, a friend bought a Nvidia PCI card instead of the AGP (old computer) 9800 or X800GTO that I offered to give him, only as he percieved the Nvidia product as being superior than the ATI. He is a NOOB, and now suffers stuttering and wishes he had the X800GTO.

wolfive started this argument before, but i still think ATi cant get ahead in single gpu territory, some say multi gpu is the future, i disagree, one powerfull gpu all the way.

so in certain arguments, the 4870X2 shouldnt even be compared against a GTX280, the only reason they are is because of pricing, and shoehorning two gpu's onto one, big, hot, card.

not that a GTX280 isnt all of those....

If you prefer one poweful gpu that's your call but I don't think it's accurate to say ATI can't get ahead with a single gpu when it's pretty clear that they don't want to. You may not like the x2s or eventual x4s etc. but ATI seems keen on the idea and so far it seems to be working for them. It's a cheaper option for a company that's still in the red that still lets them stay competetive in the high end. It'd be silly to think that they're gonna deviate from a plan that's working well just to prove they have the bigger collective penis.

But at any rate it looks like the new rumor is that Nv's next gen cards won't be out until Q4 2009 (forums.vr-zone.com/showthread.php?t=343927), so if that's true ATI will probably have the fastest single chip for a 1 Quarter+ after they release the 5k series anyways. Then we'll have to see what the GT300 is and how it fares.

this is just sad :shadedshu

they should rename 9800GT to 10800GT, then its better than HD4870 X2;)

Winterwindthey should rename 9800GT to 10800GT, then its better than HD4870 X2;)

actually i expect something like this to happen when nvidia goes through with their new rumored naming scheme

So long as nVidia doesnt relabel the 280GTX to 1400000000GTX (ie number of transistors for extra leety nerdness), any improvement will help. Problem is they are running too many different architecture SKUs to "normalise" their naming convention.

Yea the 9800GT is gonna be the GTX 150 soon here..

OMG :eek: :eek: :eek: It`s soo fast!!!!! by 100 frames different!!

ShadowFoldYea the 9800GT is gonna be the GTX 150 soon here..

I actually think I heard something like that not to long ago.

DarrenAccording to Hexus.net the 4830 is indeed faster than the 9800 GT - bearing in mind the 4830 is a new card and drivers have not yet matured yet the 4830 beats the 9800 GT!

They use the exact same architecture as the rest of the 48XX series, which is a revamp of the architecture theyve been using for a while now, i doubt drivers will see huge gains anytime soon, yet both companies have a ways to go in terms of optimisation for new games and multi gpu scaling/support.

W1zzardlook at the powercolor 4830 numbers. the performance of the 640 sp amd hd 4830 is equal to that


9800 GT is 100% identical to 8800 GT (exception being the stickers on the card)

i know its the same. my point is that this is just a ridiculous press release... the average consumer doesnst know an 8800GT and a 9800GT is the same, so when they see this they'll piss themselves laughing at the idiot who got the numbers wrong.

wolfThey use the exact same architecture as the rest of the 48XX series, which is a revamp of the architecture theyve been using for a while now, i doubt drivers will see huge gains anytime soon, yet both companies have a ways to go in terms of optimisation for new games and multi gpu scaling/support.

Wrong.


No ring bus memory controller.
Fixed the AA issues with the 3XXX series, almost like having a free AA 2X
More SP's/improved core architecture.
4870 Has DDR5



So far as driver optimisation, revisit my comment earlier, Nvidia does get driver optimisations, ususlly they kill some eye candy to get more FPS without telling the end user.

Tbh imo nvidia cant accept that in the mid range market price around the £80-170 range (nvidia playing the bullshit card) which is all the hd48XX series they just got havnt got any answers to the 4830, and 4850 no matter how meny times they try to revamp a 8800gt and Add the plus (+) give it another name to ,regardless of overclocked edition they still dont compare bang for buck for the performance ati have layed down sledgehammer and got it right. Nvidia playing stupid childish games...

SteevoWrong.

i understand your point of view, please dont post like its fact.

neither you or i truely know wether these cards will see a huge benfit from new drivers.

and ALOT of the arcitecture is the same, just optimised physically, doesnt necissarily translate into the need for new drivers.

also this card is VERY similar to the 4850, which has been out for months, any big fixes, we've already seen IMO.

uber FAIL!

.....Nvidia needs to stop hatin wow lol.

I never said that ATI has released driver tweaks spacificly, or that there was room for improvement (Far Cry 2 driver fix, before release) here recently they have focused on hardware stability and game issues, two bigger factors than if you can get another two FPS in a game or another 200 points in a benchmark.




Real gamers feel that if you can play the game at reasonable settings, its good to play. If your hardware can't perform due to a driver/BIOS bug, they would rather have a fix and it work right than a halfass driver optimisation and whoops, where are my textures, and gee that sure looks funny, damn it locked up again or some of the other BS tha Nvidia has pulled out of their ass to make fanboys gobble down the loaf they just laid.


That is fact, not opinion.

I like their use of this graph:


Yeah the 9800GT is 19% faster!!!! To bad nearly all of the cards (including the 9800GT) are getting unplayable frames at these settings.

-Indybird

SteevoI never said that ATI has released driver tweaks spacificly, or that there was room for improvement (Far Cry 2 driver fix, before release) here recently they have focused on hardware stability and game issues, two bigger factors than if you can get another two FPS in a game or another 200 points in a benchmark.




Real gamers feel that if you can play the game at reasonable settings, its good to play. If your hardware can't perform due to a driver/BIOS bug, they would rather have a fix and it work right than a halfass driver optimisation and whoops, where are my textures, and gee that sure looks funny, damn it locked up again or some of the other BS tha Nvidia has pulled out of their ass to make fanboys gobble down the loaf they just laid.


That is fact, not opinion.

actually no thats your opinion again, its how YOU feel about this situation, its not how everybody feels, thus not fact.

if you dont like them then go do something about it.

SteevoWrong.


No ring bus memory controller.
Fixed the AA issues with the 3XXX series, almost like having a free AA 2X
More SP's/improved core architecture.
4870 Has DDR5

I completely disagree with you. This post is like saying that the Phenom architecture is completely different from the Athlon architecture. Also you point out that the 4870 has GDDR5, but that argument is irrelevant to what we're talking about here.

indybirdI like their use of this graph:
. . .

Yeah the 9800GT is 19% faster!!!! To bad nearly all of the cards (including the 9800GT) are getting unplayable frames at these settings.

-Indybird

that is a prime example of a piss-poor marketing tactic nVidia has been using for what seems like ages now . . . they routinely whip out the benchmarks at the highest resolution settings, why? Because the percentage difference between their hardware, and their competitors is much higher . . . and when you're getting your ass stomped for the first time in a couple of years, you'll stoop pretty low to prove a point.


For example, in that image you quoted, they claim a 19% difference between the 4830 and the *8800* . . .

In the right-side chart, the *8800* rakes in a bench score of 19.2 FPS, the 4830 they're "competing" against only scored 16.1 FPS . . . doing the quick math, that really works out to 16.1% (that's right, the dumb-asses that prepared that slide goofed on the final figure :shadedshu) . . .

but, if you compare the differene on the left-side chart, where the *8800* scores 54.3 FPS versus the 4830's 50.3 FPS, the math now works out to 7.3% better.




What looks better to the n00bz, 19% 16% or 7%?

anyone can overclock the difference out anyway, with a bit of keen tweakage.

imperialreignWhat looks better to the n00bz, 19% 16% or 7%?

ROFL.

what this *really* looks like to me, was some beginner in marketing made this up and sent it to his boss "hey boss, we OWNED ati lol" - the boss/supervisor was like... hey, you're right. and it got posted around a bit. Add in a leak, and here we are - ridiculous, stupid slides with incorrect information.