Exactly. A Mac version is way more important than a PhysX-enabled version.

Or alt firing modes..

Well that's the thing: recent CUDA developments have opened up intense physics processing to anyone with a recent (w/i the last 2 years) graphics card from either major developer (It's been ported to ATI cards). You can download nVidia's recent demos and checked them out; runs buttery smooth on my 8800.

Will it run on my x1600 agp pro?

I bought it around 2 years ago.

I'm gonna go with no, haha.

What. They never told me about catalyst coming with that. I'm sure it'd all run better on my 4870 though.

ati does not have physx you liar I SHOULD CALL THE PO-LICE

Actually, sounds like the guy got it running on a 1950, so you might be in luck. http://gpgpu.org/forums/viewtopic.php?t=4139
Sorry, wrong source. http://www.tomshardware.com/news/nvidia ... ,5764.html

I mean, i'm just using myself as the standard user with the average specs of the forum, not just because I can use it. But if I can use it, then by all means, Data shouldn't put it way high up on the list, but putting some priority to it wouldn't be a stupid move.


Because this is basically using a component that is useless to CC and using it to process physics separately from the cpu.

So? You can't download them. I don't see your point.

By the way CC doesn't even use graphics card acceleration so PhysX is pretty much completely out of the question.

Daman, do you have to nay-say (without regard to accuracy) everything? I don't see how it's constructive to the discussion; even assumed accurate it's beside the point. If you have an nVidia card you can check out the demos right now, if you have an ATI card, it's a reasonable assumption that it will be long-successfully-ported by the time any games are using the technology. It's not as if we're expecting Data to look and this and rewrite his entire code in one evening, its just to discuss the practicality of such a move.

When build 1.0 is ready, it seems like a reasonable idea if you could gain extra performance from such a move. Maturation of development tools would also be a plus. I can't imagine it'd be too difficult to port CC's physics engine to a dedicated language, considering CC's function (physics-wise) is biased towards realism; it's likely the same calculations.