I don't expect them to be priced too comeptitively unless their performance is truely piss poor ( in terms of the delay in their release and projected performance goals )
They won't be able to do a dual gpu card on 40nm ( with a high cuda core count and clocks ) so for their sake the 480 better sit between the 5870 and 5970 comfortably otherwiseat Nvidia. Best theyll be able to muster is a dual card with substantially reduced core count as well as clock speed. What they did with the 295 was acceptable (slight reduction from a full 285, theyd have to castrate a 480 to get 2 gpus running in a 300 watt power envelope )but I wouldn't expect them to do anything like that with the GF100 until a process shrink ( GT200 to GT200b for example ) However since they are so late to the party unless the cards really impress and surprise us by the time they could realistically have a GF100b or what have you, ATI will most likely be ready with their next series making it all an even more difficult situation for Nvidia strategy wise. In summary unless GF100 is the second coming of G80, Nvidia might be in some trouble over the next few quarters.
What Nvidia would want is a 470 slightly above 5870 cost ( lets say 425-449) and a 480 at 549-599 or so. Obviously performance would have to justify it but its what I am personally expecting. Given the lack of early availability however ( all indicators so far point towards this ) if they are worth the cost theyll probaley go for even more due to the demand ( and resulting vendor mark up ) We shall see...
Bookmarks