Computer Chess Club Archives


Search

Terms

Messages

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

Author: Matt Taylor

Date: 14:28:18 03/07/03

Go up one level in this thread


On March 05, 2003 at 11:41:13, Robert Hyatt wrote:

>On March 05, 2003 at 01:23:51, Jeremiah Penery wrote:
>
>>On March 04, 2003 at 23:09:14, Robert Hyatt wrote:
>>
>>>On March 04, 2003 at 22:06:53, Jeremiah Penery wrote:
>>>
>>>>On March 04, 2003 at 00:24:27, Robert Hyatt wrote:
>>>>
>>>>>On March 03, 2003 at 22:33:57, Jeremiah Penery wrote:
>>>>>
>>>>>>On March 02, 2003 at 23:24:16, Robert Hyatt wrote:
>>>>>>
>>>>>>>And I'm going to buy the fastest thing I can at the time I purchase.  If they
>>>>>>>lag with clock speeds, I may well go with someone else.  And I believe they
>>>>>>>know that.
>>>>>>
>>>>>>Funny then, that you've never had an AMD machine, since they were faster than
>>>>>>Intel machines for quite some time.
>>>>>
>>>>>As I mentioned, we _had_ a few K5 processors.  They left a _terrible_ taste.
>>>>>I helped a Ph.D. student debug for a couple of weeks, only to find it was an
>>>>>unreliable AMD processor.  Ran fine on equivalent Intel chips.  Not on K5.
>>>>>We later find that that batch of K5's had some problems.
>>>>
>>>>I never claimed anything about the K5.  K5, by all accounts, pretty well sucked
>>>>anyway.  I'm talking about the last couple years, where Athlon was clearly
>>>>dominating performance numbers everywhere.
>>>
>>>
>>>Fool me once, shame on you.  fool me twice, shame on _me_.  Sound
>>>familiar?  That is a problem for AMD, IMHO.
>>
>>So you were 'fooled' once by a bad batch of K5s. You could have been fooled by
>>the Pentium FDIV bug, and then by the non-functional P3 1.13GHz chips.
>>In fact, over the last several years, Intel has had more problems like this than
>>AMD.
>>
>>What, exactly, is the point?
>
>Intel recalled their chips.  AMD denied they had a problem for at _least_ two
>weeks.
>
>Agreed, one bad experience, but at the loss of a month of my time, it was bad
>enough.
>
>And then there was the K6 issue with cmov missing, if you were on the crafty
>list at the
>time you'll remember all the testing and debugging everyone was trying, until it
>became
>apparent that only AMD processors were failing...
>
>That is the point.

And that was not AMD's fault as their processors were compliant with spec. I
don't know why you expected Pentium 2 binaries to run on a K6. This has never
been true of even Intel chips. (Actually I suppose it happens in applications
code, but new chips always extend the ISA, and different companies have never
been quite at the same point in ISA implementation.)

-Matt



This page took 0.01 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.