Soldato
Larabee will be a highly programmable, x86 compatible, GPGPU monster. It won't be a gaming card.
Mark my words.
Mark my words.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
nVidia GT300 - GeForce GTX 380 yields are sub-30%
- Thread starter Final8y
- Start date
Associate
- Joined
- 28 Apr 2007
- Posts
- 490
While AA performance in DX10 titles was affected (although it is hard to judge since we don't exactly have a huge amount of 10.1 games to compare to, which you can blame on nVidias irritating TWIMTBP program), it was a pretty huge oversight by ATi since pretty much all the games at the time were DX9, where the AA would have to be done in the shaders. This lead to worse performance in last gen games (by a long way), and lets face it, when a new gen of cards come out for an upcoming DX version, last gen performance is arguably just as (if not more) important than how it will perform.
My main reason for stating that it was unbalanced was due to the 512bit memory interface which provided 105gb/s~ and took up a huge amount of die area while giving virtually no performance increase over what could be achieved on a 256bit bus (as proved by the 3870 which had a 256bit bus with GDDR3 and 60-70gb/s of bandwidth, while only being a couple of % slower).
Also the ring bus did add performance, however it didn't add enough performance relative to the die space it took up (it was mainly a marketing tool really). This was also of the major reasons why in the 38XX to 48XX transition-when the ring bus was removed-we were able to see the increase in shaders and TMUs that we saw, even though it was on the same manufacturing node and the die only increased from 190mm2~ to 256mm2~.
The HD2900 never had DX10.1, so I cannot see how you are even making this comparison. The HD2900 was only capable of DX10 and DX10.1 support was introduced with the arrival of the HD3XXX series.
Soldato
Don't be stupid. No one would ever buy the card. That's hardly good business sense to sell a card for £700. Not even Nvidia would be that stupid really.
Your right.. that would be stupid. http://www.overclockers.co.uk/showproduct.php?prodid=GX-203-AS&groupid=701&catid=56&subcat=1324
Associate
mmmmm interesting why do they do this to us!!
these problems are either a load of bull, or they could be waiting ON PURPOSE to see what the 5870 does, performance wise....
because if i was Nvidia; it makes huge sense not to release anything, until you know the performance of the 5870, especially the 5870 x2........just in case, my card isn't powerful enough.
my guess is, the GTX 300 is either totally ****** up, or they're waiting on purpose....and it's already up and running, but only needs fine tuning if needed!
because if i was Nvidia; it makes huge sense not to release anything, until you know the performance of the 5870, especially the 5870 x2........just in case, my card isn't powerful enough.
my guess is, the GTX 300 is either totally ****** up, or they're waiting on purpose....and it's already up and running, but only needs fine tuning if needed!
if this card ever drops to 500 quid then i'm getting it........sod DX11, it's not needed.....you can barely tell the difference, plus i doubt many future games will be DX11............will it drop to 500 quid ?................never
Soldato
So the R600 couldn't do hardware AA in anything bar dx10.1, and that's microsofts fault?
what was the actual reason Drunkenmaster? what stopped the r600 core from doing hardware AA?
drunkenmaster said:
not with aa it wasnt, infact it was closer to the 8800gts and it consumed more power than a gtx as well.
Last edited:
I'm assuming... tho I've not actually researched it, ATI intended with the 2900 that "AA" would be applied as part of the deferred shader pipeline using an edge detection filter... rather odd approach - if thats the case I can only assume they mis-understood where MS were going with DX10.1 and thought it could be used to overload older functions to.
Soldato
if this card ever drops to 500 quid then i'm getting it........sod DX11, it's not needed.....you can barely tell the difference, plus i doubt many future games will be DX11............will it drop to 500 quid ?................never
i think it should drop to 500 easily.
im waiting for this
Associate
oooo new cards! cant wait!
Soldato
I eagerly await Nvidia's wet fart of an 'event' around October 22nd, trying to take the shine off AMD's launch.
Soldato
Apart from that stupid limited edition thing yes =p
But still. For a new release card, Nvidia wouldn't try and even sell a £500 card to a LOT of people looking for a new graphics card.
That card is £1000 because it's a limited edition, and it's so bloody stupid. Asus are simply trying to make a lot of money out of some people who want a limited edition card.
Permabanned
Thats why I said partly, not completely. After driver updates R600 wasnt really that bad either. Still below the 8800GTX but not that bad.
Caporegime
Soldato
That depends entirely on the core and shader clockspeeds.
Anyway, reading that little article does not fill me with confidence. It's very vague, and seems to be written by someone who doesn't fully understand what they're writing about. Of course it could still be genuine, but I doubt it.
I was taking that into account - just going by SPs would put it at closer to 2.7TF assuming we used the 285GTX as a reference for the other data than 3.8.