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

Re: SSE Level 3 drop in gemm



Camm,

>2) What about other compilers?  Anyone know if atlas is extensively
>   used with other compilers, and whether those even accept __asm__
>   statements?  

ATLAS is used with many other compilers, but gcc is clearly the best for
intel/linux.  In the new work I'm doing, you can specify a particular
compiler and flags to go along with your kernel, so you can insist
on using gcc, even if ATLAS were to switch to pgcc, for instance . . .

>On a separate front, I've taken the optimally generated double
>precision matmul kernel and tried adding prefetch.  Thus far, an
>inexplicable (to me at least) dramatic drop in performance,
>approaching a factor of 2!  I thought some on this list might know
>what is going on.

For what it is worth, I had the same experience on adding 3DNow prefetch
to the Athlon kernel.  I only messed with it one evening, so I never tracked
down what was going on . . .

Cheers,
Clint