I think that that the focus they put on highend chips is well placed. Highend products, particularly ones that out do the competition, garner attention. Attention is visibility, and visibility is vital to marketing. The performance crown is an important aspect of the business.
Where nVidia fails, is that they continue with this strategy throughout the life cycle of a product. Once a performance part is released, they should then focus on refining, limiting, and segmenting the arch. to fit into different markets. Go for the highend first to prove they've got a decent part and they're still relevent, then scale it back to keep sales up in all the market segments, which becomes easier once yeilds improve. Lower end parts need a great deal of supply capacity, because they're obviously going to apply to a greater market segment than the highend chips, and thus have more demand. When you're dealing with new architecture, yeilds on working chips may not be too good... As it looks like is happening with GT300. If nVidia focuses on highend parts first, where price premiums impact decision making to a lesser extent, then lower quanities and thus higher prices are much more acceptable. It's the low-mid range that would suffer horribly from inflated prices and non-existant availability. It's cool for a person to have a highend, kickass part that's only available in limited quantites. It's just frustrating to have a mid range card that you were charged a premium for because the company can't get their
together.
nVidia TRIES to do this, but it's a too little, too late situation. By the time they get around to it, the competition *cough*ATI*cough* has already extended it's arm into the other market segments. nVidia is left trying to rally its troops and get products out when the rest of the market is already into a full on invasion. Look what happened with the 8 series cards. Amazing performance for the time, compared to ATI's offerings, but they focused on the highend segment for too long. The 8800GTS and 8800GTX were the only worth while cards for a long time. Their lower end parts were overpriced for the performance. Hell, a 7600GT was still a competetive part compared to the similarly priced 8xxx series cards in those days. ATI's 2xxx series was decent enough, the 3xxx series was whatever it was, a dx10.1 update and die shrink IIRC, but they didn't really get off their asses until the 4xxx series. By then nVidia had some decent midrange products out, but the G200 cards were still in the highend stage, and the only midrange offerings they had to speak of were 8 series cards, either branded as such or under the 9 series rename, to keep up appearances that they were actually doing something with their time. That gave ATI time to catch up, grab some market share, and get products out that could compete or beat the G200 cards, all while maintaining and building a market share in the mid-low end.
nVidia has the right theory going of release highend parts first, then refine them to fit other market segments and simultaneously ramping up clocks and performance on the highend as yeilds improve, their problem is implementation. ATI has a similar, but different strategy going, where they've got their highend offerings, but from experience they seem to be throwing more thought into their mid-low end products, where nVidia's presence is lacking. The difference, is that they can actually put their plan into action. It seems like nVidia just kinda releases the highend cards and keeps hyping them to death, while lackadaisically developing their lower end offerings, which end up failing spectacularly, all while they're blind to the fact that ATI has competitive solutions available already, in all segments.
Originally Posted by RPGWiZaRD
) 128TMUs 76,8GTex/s vs. RV870 80TMUs (@850MHz) that reaches only 68GTex/s. But on the other hand do they really think they could make single chip that will make ATi crawl again without some really new radically architecture?!
(Does anybody knows what charity means these days) It's not charity just a common sense about ATi's MSRP. Stores could easily raise that prices if demand overcame supply. So it would be some common sense for ATi. But hell with reason when there's enough unreasonable buyers 
Bookmarks