🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.sys.next.misc
Article #4354

Re: 100 Mips Intel NeXT. - Performance mesurement

#4354
From: Gerhard Moeller
Date: Fri, 11 Dec 1992 07:51
49 lines
2757 bytes
In article <1992Dec8.205422.746@qb.rhein-main.de> vhs@rhein-main.de (Volker  
Herminghaus-Shirai) writes:
[...]
> MIPS for computer chips is just as meaningless as revs/min is for
> car engines, especially if they are claimed by intel.

True.

[...]
> Never trust anything before SPECint92/SPECfp92 (not that they are
> especially trustworthy...).

Not really. For example on superscalar processors (SPARC) this isn't a good
test, either.

[some good comments which I can agree with deleted...]

One thing I will never understand:
People tend to "measure" their cars in terms of horsepower and speed,
their houses in squaremeters, their girlfriends in (...) ... and
finally their computers in terms of "MIPS".

As a statistican once said: "We don't know what to measure and why --
but we do a damn good job on it!"

But "performance" and "usability" is a much more complex topic.
Terms as functionality, integrity, robustness and psychologic
impacts are MUCH more relevant to it as any speed. For example I work
MUCH more productive (objectively) on my NeXT as I do on a big MVS
system or even a DEC. (Though they are much "faster") Of course
sometimes I am waiting for my NeXT (at the moment my swapfile is about
60 MByte big) and some jobs require a higher speed (Mathematica,
rendering and some other jobs) -- but who of us really needs that?

Most of the time my system is waiting for me -- not vice versa.

You always need the right tool for the right purpose. Same with cars. A
friend drives a Porsche 928 Cabrio -- quite fast car with many HP -- but
once you had to sit in the back you start hating it. No family car. (And
in the US you couldn't even drive it fast...)

	Just my impression, Gerhard.
--
+---------------------------< principiis obsta! >---------------------------+
N  Gerhard Moeller, Teichstrasse 12, 2900 Oldenburg (FRG) [*: 02/21/1968]   N
e  Private: gemoe@proximus.north.de          Phone (voice): +49-441-75520   e
X  Uni:     Gerhard.Moeller@arbi.Informatik.Uni-Oldenburg.DE      NeXTmail  X
T  Z-Net:   Gerhard.Moeller@uniol.zer                          encouraged!  T
+-> NoGeNUG - Northern German NeXT User Group:  NoGeNUG@proximus.north.DE <-+
Technical Details
Message-ID: <1992Dec11.075151.20942@proximus.north.de>
Path: arc1-ca.pugleaf.net!sparky!uunet!zaphod.mps.ohio-state.edu!darwin.sura.net!spool.mu.edu!yale.edu!ira.uka.de!Germany.EU.net!uniol!proximus!gemoe
References: <1992Dec8.205422.746@qb.rhein-main.de>

Show Full Headers
Complete Article Headers
Newsgroups: comp.sys.next.misc
Path: arc1-ca.pugleaf.net!sparky!uunet!zaphod.mps.ohio-state.edu!darwin.sura.net!spool.mu.edu!yale.edu!ira.uka.de!Germany.EU.net!uniol!proximus!gemoe
From: gemoe@proximus.north.de (Gerhard Moeller)
Subject: Re: 100 Mips Intel NeXT. - Performance mesurement
Message-ID: <1992Dec11.075151.20942@proximus.north.de>
Sender: gemoe@proximus.north.de
Organization: Gerhard Moeller, German NeXT User Group, Oldenburg.
References: <1992Dec8.205422.746@qb.rhein-main.de>
Date: Fri, 11 Dec 1992 07:51:51 GMT
Lines: 49