On 18/11/2010 11:05, Thorsten Mühlfelder wrote:
Am Thursday 18 November 2010 10:51:47 schrieb Carlo Caione:
Ok, I hijack the discussion.
As I explained before in my previous email the idea is to recompile
armedslack with the right optimization for beaglebord (armv7 + neon +
cortex-a8 optimization)
How much speed improvement can be gained by this, e.g. at Vorbis/Flac encoding
or at compiling?
Good point. I'm not sure, but I think that support for NEON could ensure
a good speed up.
Here some pointers (not mine), about performance for different compilers
(but same parameters):
http://hardwarebug.org/2009/08/05/arm-compiler-shoot-out/
_______________________________________________
ARMedslack mailing list
[email protected]
http://lists.armedslack.org/mailman/listinfo/armedslack
