Date: Thu, 13 Feb 1997 18:45:27 +0100 (MET) From: Walter Hafner <hafner@forwiss.tu-muenchen.de> To: freebsd-questions@freebsd.org Cc: Ben Black <black@gage.com> Subject: Re: UltraSPARC and MicroSPARC vs Pentium Pro ? Message-ID: <199702131745.SAA29349@forwiss.tu-muenchen.de> In-Reply-To: <9702131713.AA25713@squid.gage.com> References: <199702120330.TAA15056@f30.hotmail.com> <199702121833.KAA18506@freefall.freebsd.org> <s9nk9ocr7po.fsf@suncog13.forwiss.tu-muenchen.de> <9702131713.AA25713@squid.gage.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Ben Black <black@gage.com> writes:
>> As you can see from the reference:
>>
>> - Overall performance is about the same as a Sparc 20
>> - float is actually much faster on the Pentium as on the Sparc 20.
> so a 133MHz P5 beats a 60MHz SuperSPARC? amazing.
Yup! And that's not just benchmarks. Our "real live" applications are
faster on the P5 too ("Real live" means mainly byte-images).
>> Considering the price, the Pentium is of course the best you can get -
>> at least for image processing! (PC's have faster and better graphic
>> boards too, compared to typical workstations!)
> yeah, those creator 3d boards with ALUs in the VRAM are just such junk. i'd
> much rather have a nice matrox board. gimme a break.
I wrote about "typical workstations". I don't consider a Ultra
200/Creator3d a typical workstation!
As a matter of fact we own several Creator (not 3d, however!)
Ultras. Unfortunately we don't have a SparcCompiler 4.0. (still waiting
for the campus licence ... :-( ). And yes: Images are considerably
faster displayed on a Millenium (Accel-X 2.1) than on a Creator (gcc-
resp. SparCompiler 3.0 compiled). Ok - the test is not fair. But I
didn't claim that. :-) I can run a xperf if you like. Now, with
optimized graphics code the results will surely look different.
Note, that I didn't try to generalize anything. I'm speaking of our
needs and observations in the field of image processing. Nothing else.
You may now say, that the benchmark on the Ultras is 3.0 compiled,
too. True. But inherently there isn't much in the code, that is Ultra
specific. I don't expect the code to be more then 10-20% faster, once we
have the 4.0 version of the compiler.
>> BTW: A P-Pro 200 has an overall benchmark of 3.0 ... faster than a Ultra
>> 143 or Indigo 2! I can't give you exact results since our P-Pro is
>> currently in San Jose (SPIE conference exhibit).
> a 200MHz P6 beats a 143MHz UltraSPARC? amazing. according to your
> benchmarks, the 167MHz UltraSPARC beats the 200MHz P6. how about numbers for
> a 200MHz Ultra?
We don't own one. But I can provide you with the benchmark, if you
like. It's only one binary. :-)
-Walter
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199702131745.SAA29349>