> > > for the other one I can`t say nothing. The main part of speeding up
> > > processors will be done be using multiple processors.
> >
> > Why do you believe there will be no 128bit processors? Isn't "more data
> > at a time" always better?
>
> Current 64-bit chips (Alpha family, I don't know about merced, but I
> think these were the patents Digital was suing Intel about, two years ago)
> do some acceleration by performing two (or more?) wholly separate
> operations at the same time, when neither operation is going to stomp on
the other.
The Merced processor which has a release date as intangible as that for NT
5, has a new method of multiple execution threads (new for Intel anyway)
generically termed explicit parallelism.
Basically, with today's Pentium line and even the 486 line, the processor
would pipeline instructions and the Pentium introduced the concept of dual
integer operations.
It was left up to the processor to try and optimize the use of this
pipelining and parallel execution which meant the processor became very
complex.
With the introduction of Merced, it is left to the programmer (the compiler)
to schedule multiple threads for execution. This makes the design of the
chip much "lighter" and gives the edge to the compiler as to how well a
program takes advantage of those features.
What this means is that, in general, a program will run no faster on a
Merced than it would on a Pentium II line, beyond any increase in MHz that
is. Only when a program is recompiled with a Merced savvy compiler will you
get any benefit from the advanced features.
I'm sure George will do that though :-)
> > I agree with you that advances in multiprocessing are needed and will be
> > beneficial, but in the race for a better desktop machine, SMP does
little
> > to help.
>
> An endless treadmill. What do you know about symetrical multiprocessing
> that makes it worthless? It is where all OS designers are heading these
> days it seems, you know.
The current trend in Intel architectures doesn't make SMP very scalable
beyond 2 processors. Beyond that, you see diminishing returns. Newer SMP
models for the Intel architecture promise to be much more scalable. The
newer XEON processors will help some, but there's still a ways to go before
it will get to the point where doubling your processors would double your
speed.
> > Without more info on the technical reasons behind your conjecture, I am
> > willing to bet that at least one of the non-intel chip manufacturers
> > announces a 128bit chip, if only to gain some attention.
>
> This too is possible, for those applications that really do need
> double-double-quad-word precission all the time.
There, basically, are 128-bit chips already, used all the time for video
cards. Would something like that make sense for a CPU? Maybe not for a
while.
