I quite agree, but that's a slightly different scenario. In this case, the Physx is being run on Nvidia's own GPU's - unless of course, someone else is doing the rendering. While there's likely some valid arguments in the programming/compatibility arena for this decision, it does smack a little of childishness - "You won't use our stuff 100%? Then take THIS - no accelerated Physx for you."
Honestly, given the right titles, I wouldn't be entirely adverse to throwing down ~$100 for a secondary card to get some sweet eye-candy at playable framerates. But dictating what the primary card has to be... meh.




Bookmarks