G80 will soon be superseded by it's refresh, so by the time R600 arrives, it looks like ATI simply did not have a competitive product while Nvidia cleaned up with with the fastest DX9 hardware, the fastest Vista hardware (even with dodgy drivers), and the halo effect that all that implies.
What I'm saying is that as regards DX10 it is apparent that nVidia does not presently own that market space--seems to me that it is presently up for grabs--at least if you use the measure of DX10 driver development and support and the number of DX10-required games currently shipping. I see no lock on that market by any IHV as of yet.
I doubt AMD is thinking that they don't need to bother just because Nvidia's current Vista drivers need improvement. Nvidia's drivers have needed loads of work for a long time, and they just shrug it off like water off a duck's back. There is still all the XP and DX9 market to consider, there is still the fact that G80 is a much faster and more advanced card with more features than anything you can buy from ATI, and that Nvidia has had free reign to both sell and market that product with impunity for nearly half a year already.
Perhaps ATi is thinking it wants a better launch for R600 than nVidia has enjoyed for G80, because ATi is indeed directly targeting DX10 and Vista support for R600 as primary goals for the R600-family product release? G80 certainly didn't stop me from running out and buying a x1950 Pro AGP product--I didn't even consider G80, actually. Heck, although I do a lot of 3d gaming, I haven't even seen the need to move to PCIe as of yet, and have yet to suffer for it to my knowledge...
I hardly think that I am alone in this regard. Some people salivate like Pavlovian dogs at the sound of the bell when they are exposed to marketing campaigns--some people don't. I'm one of those people motivated more by substance than marketing.
If Nvidia does have marketing issue with G80, they won't care. They'll expertly brush it under the carpet, and sell you the refresh, and all the while there's nothing to provide a credible alternative from AMD. AMD needs to pull it's finger out and sell us good products, instead of just talking about it. It's no use having a better card and drivers if no one can buy them, and it's no use beating the opposition roundly just as the opposition brings out a newer better product that negates yours.
The question is not primarily about what *nVidia* will do ...
The question, and what it seems to me we've been talking about, is what nVidia's *customers* will do if nVidia's current marketing problems relative to these G80 DX10/Vista support issues aren't addressed to the satisfaction of those customers, both current and potential. I think you are dead wrong about nVidia not caring about how well its marketing correlates to the products it's selling.
I think, if anything, they are very sensitive to it in light of the company's vast experience in graduating from the School of Hard Knocks relative to its lengthy and counter-productive nV3x marketing, loss of the xBox contract, and several other things I could mention. More recently, let's look at nVidia's current practice of writing bug notes with its current driver releases--when I owned my last nVidia product back in 2002--nVidia didn't "care at all" about doing that. Now, they do, and they do it. What about the recent nVidia announcement that it is moving to a monthly driver release schedule in response to the criticism it's been taking about its current G80 Vista/DX10 support? Does this indicate to you that nVidia doesn't care about marketing problems? It indicates the opposite of a "don't care" attitude, imo, and I think proves that nVidia cares about these things very much. I mean, the following imaginary scenario doesn't seem at all likely to me:
JHH: "I think we may have a marketing problem with our "Vista-Ready" G80 marketing."
Head nVidia driver developer: "Who gives a shit?"
JHH: "Yea, you're right--we'll just sweep it under the rug. Screw it--screw 'em all!"
Nope--I don't think this is at all indicative of nVidia's attitude in 2007 relative to how it markets and supports its products. And this is precisely the kind of thing ATi would certainly wish to avoid tagging along with the upcoming R600 product release.
In summary, what I'm talking about here is not whether the G80 is the most advanced DX9 product currently being sold--I think that everybody knows that it is. What I'm talking about is why it may be that ATi has pushed back the release of its R600 product line, if indeed it has actually been pushed out and if indeed ATi's internal plans have actually changed. I think that it is very likely that ATi believes that solid Vista/DX10 support for R600 are likely to be the largest motivators for buying R600 products--at least as large a motivating factor as the R600's raw performance in DX9--if not larger. As such, I think that ATi wants to do a better job with its R600 DX10/Vista support than its largest competitor has done to date, and I would find that attitude on the part of ATi to be completely unsurprising for all of the obvious reasons.
Be honest: you know as well as I do that the people who decided to buy products as expensive and as cutting edge as the "Vista-Ready" G80s, prior to the release of Vista, were certainly, somewhere in the back of their minds, considering that purchase in light of its future suitability not only for Vista but for DX10 support as well...
I think that this is the sort of PR problem that ATi would do well to avoid, and if it should cost an extra month or so, then it is still exactly what ATi should do relative to the launch of R600.
