Kurt Williams

Depends on the test case and the size of the collision objects. If you have large detailed ground collision geometry as a test case, hundreds of small convex boxes would be better than one massive concave geometry. It'll also allow you to cull out unneeded geometry for faster processing times.

keyframe

Zovreign
I've had cut and paste issues since Houdini 2.x in the Irix SGI days.

I remember everything IRIX related to be nearly flawless.

Funny what time does.

G

Indeed, other than cut and paste, everything else seemed to be rock solid, best x-windows implementation I had to use in production.

I don't missing paying $60,000 for a texture memory upgrade, no wonder Nvidia cleaned their clock. Of course back then you could charge for more stuff to off set things.

I wouldn't trust a mac in a production environment to drive a 5k resolution display in open gl with houdini, especially with an AMD graphics chipset.

No way no how. Been burned too many times by that to the point where I've switched 100% to running HP Z workstations and Z book laptops.

I love my macbooks, etc, but they're relegated to photo work and general purpose computing these days.

Agreed on the 32GB of ram, that's a minimum

As some people mentioned above, with the release of more and more 3k and 4k displays, people are bumping into display issues with GUI's on many high end applications. Not counting the need for 4x more GPU power to push a resolution 4x larger than normal, the hassle currently isn't worth it.

I can imagine Derek Zoolander trying to use Houdini on a 24" 5k display.

“WHAT IS THIS? A VOP NETWORK FOR ANTS!???”