Steve Harris wrote:
>
> Speaking for myself I have no intention of writing SSE by hand, I feel
> it's a big waste of time, and penalises people with decent (ie. non x86)
> CPU's. OTOH I plan to distrubte SSE accelerated binaries of my plugins
> generated with the intel c compiler.
Intel's compiler can generate code that autoselects best version for i586,
i686, P3 or P4 and it works very well. So you don't have to worry about what
kind of CPU the machine has. Anyway you can get significantly better
performance with writing the routines by hand.
And btw, there is no such compiler for 3DNow and AMD CPU's are still 3-10
times faster on FPU (dunno about P4 with P4 code). My handwritten Enh3DNow
code is still 10 times faster on AMD Athlon (Thunderbird) than same code
written in C optimized by Intel's compiler running on P3 (Tualatin) at same
clock frequency.
- Jussi Laako
--
PGP key fingerprint: 161D 6FED 6A92 39E2 EB5B 39DD A4DE 63EB C216 1E4B
Available at PGP keyservers
