Re: MISC personal computers
- To: "dirnfir" <dirnfir@xxxxxxx>, "MISC list" <MISC>
- Subject: Re: MISC personal computers
- From: "vic plichota" <atsvap@xxxxxxxxxxx>
- Date: Sun, 7 Feb 1999 11:00:20 -0500
dirnfir wrote:
>Does anyone think that a personal computer with a
>MISC CPU would become popular?
Certainly not in the sense of mass-marketing; the sheer momentum and
fear-based tactics of Windows(tm) has already resurrected the i86
beyond all common sense -- the ISA PC was dying out, with many fine
RISC architectures ready to take over, when paradoxically (and
completely against all common sense) Win95 steam-rollered the
marketplace, and took the i86 to grotesque new levels of silicon.
Remember when you could give your old PC to a kid? Not so anymore;
with the multi-media CD-ROM games, etc., kids are THE most demanding
'power-users' in the world today.
"Computer Literacy" is nowadays popularly regarded to be the ability
to run app's in the workplace -- it has nothing to do with
understanding the hardware, or being able to program.
However, there is a growing backlash of computing *professionals* who
are sick + tired of the lies and confusion surrounding MicroSoft's
products, and would just love a cheap alternative platform that they
could actually control.
>could MISC one day supplant the PC as the
>dominant computer in the home?
Certainly in MY home... right now my only real option is to run
LINUX! A MISC PC would *have to* offer open source-code, however,
before I would even consider it.
Keeping the power consumption low is a most attractive feature; a
portable system that spent most the battery juice on the LCD
backlight would be a real asset... it's astonishing how much can be
done with so little!
If millions of transistors are to be spent, then OK: spend them on a
floating-point or graphics coprocessor -- they aren't neccessary in
the CPU core.
But maybe an extremely-high-performance machine could be built
cheaply by merely adding more modest MISC units: one to handle human
input, one for the display, one for floating-point, one for
mass-storage, etc... don't laugh, this approach may be viable, and
would offer an unprecedented level of true parallelism, as opposed to
the joke of time-slicing a Pentium II.
cheers - vic