[rescue] pursuing a VAX

Joshua D Boyd jdboyd at cs.millersville.edu
Mon Dec 9 15:30:06 CST 2002


On Mon, Dec 09, 2002 at 03:27:40PM -0600, Scott Newell wrote:
> >> things like, "I've got a dual 1GHz Pentium III system, so it's
> >> twice as fast as a 1GHZ Pentium III!" which isn't close to the
> >> truth.  You can't correct them, either.
> >
> >How much faster, on average, would such a system be?  I know it's not
> >linear, but is there a good rule of thumb for that?
> 
> I've got a dual P3-450 at the house, and I used to have a single P3-450 at
> work.  With interactive use (edit/compile/debug) under win2k, the dual
> machine _felt_ maybe 50-75% faster, because the UI would remain responsive
> under load.

Did you try recompiling software for SSE?  Reportedly gcc can get a
significant speed bump when using SSE, without any other adjustments to
the code.  Of course, it depends on what the code is doing, etc, yada
yada yada, and so on.

-- 
Joshua D. Boyd



More information about the rescue mailing list