Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 12 Aug 1997 09:24:27 -0400 (EDT)
From:      "John T. Farmer" <jfarmer@sabre.goldsword.com>
To:        hardware@FreeBSD.ORG, stephane@e2c.com
Cc:        jfarmer@goldsword.com
Subject:   Re: K6-200 Has anyone successfully done a 'make world' ?
Message-ID:  <199708121324.JAA05886@sabre.goldsword.com>

next in thread | raw e-mail | index | archive | help

On Tue, 12 Aug 1997 10:38:47 +0200 (Stephane Legrand) said:
>>>   3.) Does anybody have any clues as to what excatly is dieing in the
>>>       "make world" process?
>>
>>   When CPUs are non-idle for extended periods (such as what occurs during
>>a "make world"), they get hot. My K6 appears to have developed a sensitivity
>>to heat and giant heat sinks with thermal compound can't keep it cool enough
>>to be reliable. It dies anywhere from 1 hour to 2.25 hours into the make
>>world, depending on the quality of the heat sink, how warm the room is at
>>the time, and the phase of the moon. Dies in this case means that a process
>>exited with a a weird syntax error or a sig-11.
>>
>
>So, the problem with the K6 would be caused by heat problem when the CPU
>are heavily load during a long time ?
>
>Simply a suggestion, but may be a test like several parallel intensive POV
>calculations (or others intensive CPU operations) during some hours should
>lead to the same errors and should definitely show that the problem come
>from the CPU and not from FreeBSD, no ?
>
>Just my 0.02$ :)
>

This is what I was getting at.  So far, everybody has had a problem with 
doing "make world" after some period of time.  To get AMD to pay attention,
they need to see that it's a problem that's _not_ tied to any specific
action in the make world  itself, but something within the chip or
it's operating environment.

If it is a heat related problem with boards that are supposed to be
approved for use with the K6, well, we may get the "It works fine under
NT & 95..." argument.  Ie, the average user doesn't impose the same
load on their cpu that we're talking about.  Sigh.  I really like using
the K5 chips in my servers & was looking forward to the performance of
the K6.  If it can't run load under 7x24 conditions, then it's _not_
an option for me.

Anybody know how the Cyrix chips hold up under heavy server usage conditions?

John	(Who doesn't want to go back to "Intel Inside"...)

-------------------------------------------------------------------------
John T. Farmer			Proprietor, GoldSword Systems
jfarmer@goldsword.com		Public Internet Access in East Tennessee
dial-in (423)470-9953		for info, e-mail to info@goldsword.com
	Network Design, Internet Services & Servers, Consulting



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199708121324.JAA05886>