Date: Wed, 12 Feb 1997 10:33:31 -0800 (PST) From: "Jonathan M. Bresler" <jmb> To: mcwong@hotmail.com (M.C Wong) Cc: questions@FreeBSD.ORG Subject: Re: UltraSPARC and MicroSPARC vs Pentium Pro ? Message-ID: <199702121833.KAA18506@freefall.freebsd.org> In-Reply-To: <199702120330.TAA15056@f30.hotmail.com> from "M.C Wong" at Feb 11, 97 07:30:14 pm
next in thread | previous in thread | raw e-mail | index | archive | help
M.C Wong wrote:
>
>
> Hi there,
>
> I am making recommendation for using FreeBSD or BSDI as a WWW server
> for a hospital WWW server and I have the following on the prefered
> list:
www server is all integer code, not floating point.
(bear with me, the distinction is important)
>
> Sun Netra i/170L (UltraSPARC 167MHz) as WWW, SMTP, caching proxy server.
> Sun Netra i5 (MicroSPARC 110Mhz) as firewall.
>
> On the other hand, I am keen on recommending FreeBSD or BSDI with
> the following hardware:
>
> Intel Pentium Pro 200 for WWW, SMTP, caching proxy server and
> Intel Pentium 166 for firewall.
>
> However, I need more real-world benchmark for the following CPUs:
>
> UltraSPARC 167MHz vs PPro 200Mhz, and
> MicroSPARC 110Mhz vs Pentium 166Mhz.
>
> Without some hard figures showing comparison, my recommendation will
> not be too convincing. Can anyone help ?
i dont have a www server benchmark numbers available,
but i do have results for an excellent cpu/cache/memory
benchmark called "Hint".
quick dirty answer:
Integer: the intel boxes kill the snot out of *all* suns
Float: the ultras outperform intel boxes.
long answer:
the sparc architecture is limited in its ability to perform
integer operations. my suspicion is that the memory bandwidth
is not up to the task. (surely, its not the cpu itself, but
rather feeding data and instructions to the cpu that is the
limiting factor.)
some performance ratio: (re 586-90)
integer:
cpu data set size
10kB - 1MB
ross 125: 65% - 90% of a 586-90 (yes less)
ultra 167: 55% - 80%
sparc 20: 40% - 60%
ppro 200: 350% - 400%
ppro 150: 250% - 300%
get the Hint benchmark and hammer some systems.
read the paper to appreciate the work that these guys
have done for everyone.
http://www.scl.ameslab.gov/scl/HINT/HINT.html
note: the interactive graphing tool uses floating point
data, not integer. (these guys are doing finite
element analysis and the like.) so the number that
you see will be different (as i said above)
jmb
ps. some data is available on freefall ~jmb/Hint.Results.tar.gz
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199702121833.KAA18506>