Re: A.M.D. to Split Into Two Operations

On Oct 13, 12:13=A0pm, Del Cecchi <delcecchinospamoftheno...@[EMAIL PROTECTED]
>
wrote:
>
> I think attributing any anti-competitive motives is incorrect. =A0Over
th=
e
> course of many years with IBM I saw many stories in the trade press
> about IBM decisions and how they happened and what they meant that were
> totally incorrect. =A0I would venture that any "kill Itanium" stories
are
> too, if only because IBM had already had a flirtation with VLIW and knew
> the story.
>
As you have heard me say elsewhere, I've heard many people say, "I
knew it all along."  Some of the objections I understand (there's too
much distance between the compiler and run-time, even with profiling,
for example).  Some of the objections may or may not be based on deep
insight, but there's never enough hint of where the deep insight comes
from for me to take the claim seriously.  I'm a long way from knowing
the whole story of IBM's involvement with VLIW, and I suspect that
most public commentators are, too.

However bad VLIW or Itanium or EPIC may be, it works well enough that
IBM bought out Platform Solutions, Inc. after using every other means
at its disposal to kill it off.  Actions speak much louder than words,
and especially much louder than retrospective analysis.

> You are upset that the Sheeple stick with =A0x86 even though it is
clearl=
y
> obvious to you that it is the wrong thing. =A0And you are disappointed
in
> our country and the free market system for allowing it to happen.
=A0"Why
> can't those fools see the correct answer to our problems"? =A0I would
> guess that most folks in this group have felt the same way at some time.
>
There is no need for you to sharpen my prose.  If I want to be
sarcastic, I know how to do it without your help.  There are powerful
economic incentives behind x86, just like there are powerful economic
incentives behind the Interstate Highway System and the ecosystem of
dependency on foreign oil that comes with it.  Markets don't magically
know how to fix problems like that.  Well, they may, but they don't
get my vote of confidence.

> So, if you are not going to let the market decide the direction of
> future microprocessor development, then who should? =A0The
undersecretary
> of commerce for microprocessors? =A0The House subcommittee on
> Microelectronics? =A0A vote among leading academics and other experts? =
=A0Or
> maybe a vote by everyone after they are shown the alternatives? =A0Only
a
> dollar a vote.
>
I'll be happy just to trash markets and those who wor****p them.  The
first commandment isn't first for no reason at all.  Superstitious
belief in the power of free markets is so pervasive that I feel well
within my rights telling people, "You are all possessed."

Robert.