--- [EMAIL PROTECTED] wrote:
> I don't speak on generic terme, i speak about real world example of
> current complexe vector shader and pixel shader.
> 
I speak of the actual code which we intend to use; it contains a lot of 
vector and matrix operations.
---
We aren't trying to make a programmable shader with these SIMD units, are
we? I think we can easily use the SIMD units in a pseudo-fixed pipeline.

Look at some previous posts on the subject:
http://article.gmane.org/gmane.comp.graphics.opengraphics/2445
http://article.gmane.org/gmane.comp.graphics.opengraphics/2553
http://article.gmane.org/gmane.comp.graphics.opengraphics/2461
These are two shader examples and an analysis of various functions used.

Just looking at that list, it is almost all dot products. Taking a previous
quote from Nicolas:
> All of this show me that GPU are almost like normal CPU that manage more
> complexe DATA (vect4 rather than int).

Looking at those posts, I think that vector operations are used in a GPU
(if we ignore the whole shader issue). Why not use the very efficient SIMD
units that have already been designed by Sun? This does _not_ mean using the
whole FPU, CPU, or even any more of the design -- I just want the SIMD unit
architecture.

I perused the ogsim code and it appears to be filled with vector operations --
there is no getting around that. I honestly don't care if there aren't even
vector instructions for programmable shaders -- if we can exploit vector ops
to make our pipeline faster, I think we should do it.
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)

Reply via email to