Date: Wed, 12 Feb 1997 18:15:39 -0800 (PST) From: "Jonathan M. Bresler" <jmb> To: jmb@freefall.freebsd.org (Jonathan M. Bresler) Cc: black@gage.com, mcwong@hotmail.com, questions@freebsd.org Subject: Re: UltraSPARC and MicroSPARC vs Pentium Pro ? Message-ID: <199702130215.SAA11767@freefall.freebsd.org> In-Reply-To: <199702130123.RAA08659@freefall.freebsd.org> from "Jonathan M. Bresler" at Feb 12, 97 05:23:33 pm
next in thread | previous in thread | raw e-mail | index | archive | help
it occurs to me that i be speaking harshly on this matter.
if so let me apologize here and now.
in my opinion, pc's are a remarkable value
they are serveral times more expensive than pc's of better
or comparable performance (however, you must pc your pc
from someone that knwos what he is about, suck as rod
grimes.)
as i understand cpu/cache/memory benchmarks, both Hint and
SPEC, pc outperform suns when executing integer code by a
healthy margin. (this is over general...586-90 is close
to a sparc20)
applications that i have run and timed on both machines
bear out the benchmark results. an example: we have an
economist at work that runs large mathematica jobs. this
code is floating point intensive. on a quiet sparc20 the
calcuation took 410 seconds, and on the 586-90 460 seconds.
yes. the sparc20 was faster. it had 256MB vs 48MB.
it cost far more. (oh...mathematica 3.0 sparc vs mathematica 3.0
linux running in binary emulation on FreeBSD 2.2)
this example plays to sun's strong suit vs pc's: floating point.
now if you are doing Java, use a sun. our java is slow, no
fault of jeffry hsu.
otherwise, consider carefully, run your application on
both platforms and then decide.
jmb
Jonathan M. Bresler wrote:
>
> Ben Black wrote:
> >
> [snip]
> >
> > as for "killing the snot out of Sun" in integer performance, the SPECint95
^^^^^^^
that should hvae been "kicking"
> > numbers just don't show it:
> >
> > SPECint95 SPECfp95
> > Intel Alder 200MHz P6 8.09 6.75
> > Sun Ultra 2 1200 200MHz 7.72 11.1
> >
> > if i get a chance, i will run the HiNT benchmarks on an UltraSPARC and a P6
> > here. i doubt i will see any snot flying from the Sun.
> >
> > benchmarks, of course, can always say what you want them to.
>
> Ben,
> please read the paper, run teh benchmarks, understand why spec
> is dubious. then lets talk about it ;)
> jmb
>
> ps. how many processors in that Sun Ultra 2 1200 ??
> i dont have a sun catalog/price list here at home.
>
>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199702130215.SAA11767>