Nvidia PDF about why more is less.
Hey guys, I've been lurking for a bit, and I decided to finally post.
Not too long ago, I was reading a post on the subject of low polygon modeling, and someone linked to a PDF from Nvidia that was explaining how it's possible to make a model render more cheaply if you add more polygons to the edges or something, because it would lower the number of draw calls that needed to be made to create the object. Something to do with the way that polygons are banded together before being rendered.
I tried searching, but I can't think of any search terms that would really narrow it down.
Does anybody have any idea what I'm talking about? or have any link to this PDF
Not too long ago, I was reading a post on the subject of low polygon modeling, and someone linked to a PDF from Nvidia that was explaining how it's possible to make a model render more cheaply if you add more polygons to the edges or something, because it would lower the number of draw calls that needed to be made to create the object. Something to do with the way that polygons are banded together before being rendered.
I tried searching, but I can't think of any search terms that would really narrow it down.
Does anybody have any idea what I'm talking about? or have any link to this PDF
Replies
these aren't pdfs you're looking for, but links to information regarding this subject too.
Beautiful, Yet Friendly Part 1: Stop Hitting the Bottleneck
Beautiful, Yet Friendly Part 2: Maximizing Efficiency
and the thread
Too much optimisation (do polygon counts really matter?)