GA is talking of 2D not 3D and talking of casual systems, not the unrealistic target systems you target with your engines Josh ... Not all have dual to quad cores with high end engines. Most here actually target the more real indie - casual market to really sell an end product, not middle ware engines where you can afford to "ignore" that little fact as this won't be your problem in the end.
So fillrate is worlds earlier a problem than you expect them to be as onboard cards have a very low maximum fillrate.
And if it would be negligible, I don't think GA would "waste" time on improving it, don't you?
I agree on the alpha point but if you have alpha creation variance this means "ignorance of the preset lifetime and additional calculation on per particle basis" to adjust the lifetime. quite cpu intense if we assume target systems of 1000mhz single core P3 and 1000-2000 particles.
So fillrate is worlds earlier a problem than you expect them to be as onboard cards have a very low maximum fillrate.
And if it would be negligible, I don't think GA would "waste" time on improving it, don't you?
I agree on the alpha point but if you have alpha creation variance this means "ignorance of the preset lifetime and additional calculation on per particle basis" to adjust the lifetime. quite cpu intense if we assume target systems of 1000mhz single core P3 and 1000-2000 particles.