Whatever happened to S-Buffers? (by alaric)
Back when I were a lad, I went through the obligatory phase of wanting to write a 3D computer game, and during my research, I came across S Buffering, a technique for rendering 3D graphics much more efficiently than a Z Buffer, the technique used in modern 3D graphics cards.
In software, an S buffer certainly seems to be inherently faster than a Z buffer, and at the time of the original article on S buffers, a software S buffer would (he claimed) outperform a hardware Z buffer.
Plus, they very elegantly handle transparency, and transparent surfaces abound in modern games.
But what happened to them? I thought about them while unable to sleep last night, so did some googling today, and all I found apart from copies of the original article was a forum post in which somebody basically asks the same question.
So I'd be keen to see if an S buffer on a modern CPU would outperform a 3D video card (I suspect video bandwidth would be the limiting factor, though). But mostly, I'd be keen to think about whether one could implement an S buffer in hardware that would produce better price/performance than a Z buffer...