Computer Chess Club Archives




Subject: Re: Introducing "No-Moore's Law"

Author: Robert Hyatt

Date: 07:22:07 03/02/03

Go up one level in this thread

On March 02, 2003 at 02:39:08, Jeremiah Penery wrote:

>On March 01, 2003 at 20:17:14, Robert Hyatt wrote:
>>On March 01, 2003 at 11:48:50, Jeremiah Penery wrote:
>>>On March 01, 2003 at 10:11:13, Robert Hyatt wrote:
>>Sun doesn't sell more than a single million-dollar computer a year.  They
>I couldn't find specific numbers on this, but I really doubt that.  How else
>could they get to $5 billion in SPARC _server_ sales last year?  Do you have any
>actual data you can give?  I'd be interested to see it.

I'll look.  Several "trade journals" quote such market shares frequently, I'll
see what I can dig up.

>>are hopeless in that market, because that is SGI Challenge territory and the
>I'd say SGI sells less big boxes than SPARC these days.  SGI is poised to have
>_total_ revenue of only $1 billion in fiscal 2003.  That's counting the Itanium
>machines they will begin selling soon.
>>SGI eats the sparc in any benchmark ever created...  As will the big
>Care to back that up with some actual data?  I can't find a single benchmark
>where a MIPS chip is above the newest SPARC chips.  Nor can I find a MIPS-based
>machine faster than any SPARC machine.

I'm not sure where you are looking.  We just bought a batch of (I think) 900mhz
ultra-sparcs (6).  They are 1/4 the raw computing speed of our best intel box
and I am not talking about SMP, just raw CPU power.  I can post a crafty bench
if you want to compare.  They are _dog_ slow.

>>itanium boxes.  Hell, even a multi-xeon will eat any sparc on a cpu for cpu
>>basis without even using SMT.  :)
>I believe you there.  But as you pointed out before, processor speed doesn't
>matter much in the server world.

No, but when you start designing high-performance busses, multiple I/O channels,
ultra-high memory bandwidth, it is likely that top-end processor chips will be
used as well, particularly 64 bit...

>>>>But that entire segment is 5% or less of the total sales.  Probably less as
>>>>mainframes are in that 5% category as well...
>>>If you sell 5 systems for $5m each, or 500 systems for $1000 each, which is
>>>making more money, despite being only 1% of the total volume?
>>The problem is, it is 5 for $5m, or 5M for $1000.  The math is _simple_
>Then the one company has only .0001% of the total sales.  I gave numbers
>representing 1% of the total sales, since you think 5% was too much.  Give it
>even 0.1%, and the revenue is still equal in that example.
>>>I said they planned to release it for the desktop - the DEC machines, not the
>>>Polywell ones - not that they ever actually did so.  I visited New Mexico Tech
>>>in 1997 or so, and they had 533MHz Alpha desktop machines there.  It seemed to
>>>me that they were trying to make that kind of machine standard equipment for
>>>students and such.
>>I have one in my lab (an alphastation from dec.)  But it was never near the
>>PC price point, which left it for the "number crunchers" where it was the best
>It never reached that price point because it never had the economy of scale.  If
>everyone had bought one of those Alpha machines, price would have gone down
>significantly (development and initial manufacturing cost amortized over a far
>greater volume of processors), inspiring even more people to buy them.  It
>failed to reach that critical mass.

Yes, because it was too expensive compared to a bare-bones PC...

This page took 0.02 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.