On Thursday 10 February 2005 15:42, Timothy Miller wrote:
> Oh, you want to move geometry perspecitve divides into the chip.

Compute 1/W for three vertices on the host and send it to the card with 
other parameters in 8.8, as a "small triangle" optimization.

> Well, maybe we can think about that, but keep in mind that the CPU is
> going to be much faster at that.  It may not be enough to have a real
> benefit.

In return for eating a multiplier it would save a lot of PCI bandwidth, 
which looks like a major bottleneck.  For what it's worth, there's a 
slight CPU offload benefit as well.

Regards,

Daniel
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)

Reply via email to