On Mon, 24 Apr 2000, Henk Stokhorst wrote:
> Hi,
>
> The new IA-32 processor under development, codenamed 'Willamette', has
a 64 bit FPU
> and an ALU running at twice the clockfrequency. However, the latencies
are different
> from the existing Pentiums.
>
> Did anybody have a look at the processor and determined if the ALU or
the FPU of the
> Willamette should be preferred for LL testing?
Preliminary results seem to indicate that comapartively speaking, integer
multiplies take *longer* on Willamette than on previous Intel processors.
The big difference here is that "Willie" will have a quad 64-bit bus to
memory, 128-byte cache lines(I think), and SSE extensions that allow for
full 53-bit double precision floating point from SSE, i.e. working on
pairs of doubles at a time. I think that there will also be SSE extensions
to do 32-bit integer multiplies, meaning that the vector stuff can finally
be effectively used for crypto and multiprecision integer arithmetic.
jasonp
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
