By: redpriest (redpriest.delete@this.nospamalum.wpi.edu), November 6, 2006 6:30 pm
Room: Moderated Discussions
But congratulations for being unnecessarily caustic without any supporting evidence. Have you seen the actual MPF presentation, by chance?
Michael S (already5chosen@yahoo.com) on 11/6/06 wrote:
---------------------------
>"support for unaligned load ops was added" probably means that K8L handles muvups/movupd/movdqu
>via DirectPath instead of VectorPath (a.k.a. microcode). That's was always a case
>on P6/PM. So in that regard AMD finally catching up with 8 years old Intel product.
>That's fine but hardly means that we shouldn't align.
Michael S (already5chosen@yahoo.com) on 11/6/06 wrote:
---------------------------
>"support for unaligned load ops was added" probably means that K8L handles muvups/movupd/movdqu
>via DirectPath instead of VectorPath (a.k.a. microcode). That's was always a case
>on P6/PM. So in that regard AMD finally catching up with 8 years old Intel product.
>That's fine but hardly means that we shouldn't align.
