[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: OpenGL vs. SDL for 2D

> I don't think you'd get much improvement using 3D accelerators for what
> be done using SIMD instructions, as long as your pack your datatype
> efficiently, and if the alpha-blending is all you need from the
> I wouldn't bother using one. It cuts out a portion of the market, see?

I know what you are talking about -- SIMD -- the idea you apply same
instruction to a large chunk of data, but still, would that compare to a
decent 3D card with hardware OpenGL? If even only a 20% perf. boost is
possible using 3D hw for the task.

> AFAIK, GCC doesn't auto-vectorize (yet - I met the guy who was being paid
> SuSE to do it), and the Linux version of VectorC isn't out yet, so you'll
> need to do it manually, or nab the source from the Codeplay website
> (http://www.codeplay.com) for the demo (or use the benchmark snippets) as
> reference.

I know, I know. Hope they make it.

> IIRC, SIMD operations for blending textures is still faster than using the
> gfx cards, as long as your data is in a decent format.

But if I can compute something on the main CPU at the same time, it would
still be faster?

We must not forget that SIMD are available only on recent processors, so
this would be leaving out a part of the market, too.

In general, what is SIMD good for in games (if we would not deal with FP
Oh well, there's so much I have to learn to know *everything* -- if it's not
possible, we
can at least try, right?