> > Accessing textures from host memory can be very inefficient. But it
>
> I just cant help it but i have to ask...
>
> how can a system designer make it efficient?
> hypothetically...
- More graphics memory so you can hold all your textures
- A faster bus between the GPU and the host
- A unified GPU/CPU/memory architecture
I think that what AMD (and possibly Intel) are doing next. Putting a
few GPU cores along side with the CPU cores all in one chip. That's
probably the main reason for AMD buying ATI.
Andy
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
