Hi, Yann Guidon wrote: > hello, > > Andre Pouliot wrote: >> Another way to view it is turn the data from a SIMD on the side. You >> change any lateral dependency of the data, you bring it back >> vertically. On a standard SIMD you would process different threads >> even with the same instruction one after the other. The shader we >> propose, we process all the same threads at once but we break down >> the instruction in more but simpler instruction. > Be sure to prepare a few illustrations to explain your ideas, > so you won't have to use long paragraphs that try to describe it :-) > > yg
I'll try to add more illustration. Having a white board to do the explanation would also help a lot. ;) I know that this processor model is hard to grasp. It took me quite a while to understand it. Usually, I'm good at that stuff. Still if people ask me question on how it work and why. That mean either the explanation in the doc is still not good enough or we were wrong to use it. I have read good question and query for explanation on how we handle some case, but still no good argument against. The argument everyone else is doing it another way isn't good enough for me. I like proof of the why. _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
