0
0

Altivec is incredibly faster than MMX on on brand new PPC9700 (Apple G5) there are 2 Altivec units.
can’t you add support for this technology ?

  • You must to post comments
0
0

That would be a lot of work to support a really small section of the market, wouldn’t it?

  • You must to post comments
0
0

I know… it’s always the same problem… it is easier, faster, and simpler as it doesn’t need any inline assembler coding… but the world is enslaved to the slower intel compatible processors

  • You must to post comments
0
0

That’s right, but it can be at least two time faster !¡!
I am not asking you to rewrite the code to setup all the momory 16byte aligned, but there are a lot of function that can be optimized

  • You must to post comments
0
0

I belive you, and in fact I don’t know much about audio programming but altivec allow you to work on 16 byte of data faster then a processor word. multiply 4 int costs 1 cycle…and with 32 128bit vector register anthing can be done.
but i think you don’t mind this, thatnks for anserwing.
bye

  • You must to post comments
Showing 4 results
Your Answer

Please first to submit.