One thing to keep in mind about a K6, is that, from what I've been told, AMD
releases their chips clocke for the maximum they'll run at in a stable
manner. Intel, on the other hand, apparently makes a bunch of chips at one
speed and "downgrades" the clock setting on some to sell them as "slower"
processors, which is why you can overclock 'em better than you can an AMD.
i.e With an AMD you get EXACTLY what you pay for, and not much more. With an
Intel processor, you may or may not be getting a chip that will run ONLY at
the specified speed -- in fact, it may run faster, but you take your
processor life in your hands if you overclock and it just happens to be one
of the few that won't run well at a higher clock speed.
John
----- Original Message -----
From: <[EMAIL PROTECTED]>
To: Daniel Landau <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Sunday, June 06, 1999 5:13 AM
Subject: Re: Signal 11 error - gone with RHv6.0 !
> According to Daniel Landau: While burning my CPU.
> >
> > Recently, [EMAIL PROTECTED] wrote:
> >
> > > signal 11 is a memory problem, it has many causes, most of which can
> > > be caused by an overclocked CPU, other reasons are _not_ having a swap
> > > space active, mixtures of memory good and bad, different speeds, ie
one
> > > bank of simms at 80ns and another at 70ns, wrong cache settings in the
> > > bios, sig11 is a big problem on K6 CPU's, to gain more information
take a
> > > look at;
> > >
> > > http://www.bitwizard.nl/sig11
> >
> > Thanks for the tip on this url. Lots of good signal 11 info there.
> >
> > One thing I would like to clarify, though. The Sig11 FAQ at that page
> > indicates that there are sig11 problems specifically on AMD K6/2-300's
> > (1998, produced in weeks 34 through 39). Your post kind of implies that
> > there are sig11 problems with all K6 CPU's. As a big AMD fan, I
wouldn't
> > want anyone to get that impression.
>
> No i'm not implying anything, only that Sig11 seems to be reported with K6
> CPU's more than any other, the reason is mostly an overclocked CPU.
>
> The linux-kernel has many K6 work arounds coded in to eliminate most (if
not
> all) off the software problems, if the user overclocks, then problems are
> bound to happen. That is what i mean, and why i quoted the above URL.
>
>
> >
> > Regards,
> >
> > Daniel
> >
>
>
> --
> Regards Richard.
> [EMAIL PROTECTED]
>
