For those lamenting that gameplay is more important then graphics- there is a PC Gaming forum, I post there all the time, it really does work. If all you care about is gameplay, it's a good forum for you. This is the Video forum, coming in here and saying graphics don't matter seems an awful lot like thread crapping. I'm not saying I disagree with you in any way shape or form, but games are allowed to have good graphics
and good gameplay. Honestly.
I see this as two different issues rolled in to one honestly.
First off, there is no big leap coming for graphics rendering on the actual rendering side for many years. We are at the point now where the 'big jumps' are going to be compression of the complexity of shaders utilized and the speed in which they run at. Outside of lighting models way beyond anything on the horizon(several orders of magnitude beyond RTRT) there isn't going to be another shocker in terms of graphics fidelity.
More advanced physics in games, particularly particle effects, are going to be the, by far, largest improvement we are going to see in game visuals for a long time now. Yes, there are other areas we are going to see progress being made, but none of them have the potential to have as much impact in a short span of time as significantly more advanced physics.
With that said, the only physics platform right now that is a viable alternative to that end is PhysX. Yes, I know that Havok will have GPU support at some date in the future. After OpenCL is finalized, after we have drivers that support it, after Havok gets ported over THEN developers can START taking advantage of it. Don't get me wrong, any game coming out in a timeframe to use GPU accelerated Havok have at it, but I was really hoping that we would see a bit better adoption of technology that can help push the industry forward. We have had the technology available to us for years already, we have an installed base of over 100 million PCs and Macs that can run PhysX right now, a bit better adoption rate from developers would be welcome(this year looks to be much better in this regard, we actually have a decent selection of titles hitting with proper support).
In terms of PhysX itself, as a physics API it really isn't much different from Havok. The only big difference at this point in time is that PhysX can run on MUCH faster hardware and hence, do more to enhance the visual experience in games. For that reason, and no other, right now I like PhysX quite a bit more then the alternatives.
How much of a factor would it be for me buying hardware right now? Well, with close to parity between parts I would rather have the part that supported it then not, and with price and performance being so close it could end up being the deciding factor for me, but that is mainly due to the fact that everything else is so close. If we were seeing faster adoption of more advanced physics by game makers, that would change things considerably. Losing 10% raw framerate for a huge visual benefit in a decent selection of games would likely be a worth while trade off for me.
This would have been so much easier if ATi would have gotten PhysX up and running on their hardware when nV offered. Then everyone could be benfitting now instead of having the loyalists from either camp fighting about it