Anyone know how to figure out what a card's vertice per second calculation speed is? Like how fillrate is calculated with the core clockspeed times the number of pipelines (and texture units in some cases). Or is there no formula for that?
Generally, vertex shader ops are equal to the core clock times the number of vertex shaders.
If you're talking about just fixed-function processing, it's not real easy to calculate the number of vertices that can be handled per second. Much of it depends on how effective the HSR functionality of the card is, and how much of the geometry is actually viewable.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.