MathGroup Archive 1994

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Alternative to Timing[10000!] benchmark

  • To: mathgroup at
  • Subject: Re: Alternative to Timing[10000!] benchmark
  • From: rubin at (Paul A. Rubin)
  • Date: Tue, 5 Jul 94 15:57:04 EDT

>Instead of the rather dodgy "Timing[10000!]" benchmark for performance,
>the following seems a better way of measuring the elapsed time taken to do a
>calculation (as Timing[] supposedly measures CPU time):
> In[1]:= start=AbsoluteTime[]; Print[Short[10000!]]; AbsoluteTime[]-start
>For the Windows 387 Version 2.2 running on a Compaq Despro 433i PC, I get the
> Out[1]= 137
>And on the DEC Alpha OSF/1 Version 2.2 version, running on a DEC 3000 Model
>400, I get
> Out[1]= 13.59984
>The "Short[10000!]" means that the timing isn't affected by the time taken
>to print 36000 digits to your VDU.
>What results to Mathgroupers get with this on their machines?
>Paul E Howland
>Long Range Radar Sensors Section                         tel. + 44 684 895767
>CSS2 Division, Room BY209                                fax. + 44 684 896315
>Defence Research Agency                           email: PEHOWLAND at DRA.HMG.GB
>Malvern, Worcs, WR14 3PS, UK.

On a Gateway 2000 P5-90 (90 MHz Pentium), 16 MB RAM, running Mma 2.2.2 under
Windows for Workgroups 3.11, I got 29 seconds the first try, 22 every try
thereafter.  (Mma was the only open window, which can make a difference, as
can the amount of RAM.)  I've noticed that many expressions in Mma evaluate
faster on the second and subsequent go-arounds, not just those that store a
result (such as f[x_] := f[x] = ...), so I always take multiple shots at
timing measurements.

Paul Rubin

* Paul A. Rubin                                  Phone: (517) 336-3509   *
* Department of Management                       Fax:   (517) 336-1111   *
* Eli Broad Graduate School of Management        Net:   RUBIN at MSU.EDU    *
* Michigan State University                                              *
* East Lansing, MI  48824-1122  (USA)                                    *
Mathematicians are like Frenchmen:  whenever you say something to them,
they translate it into their own language, and at once it is something
entirely different.                                    J. W. v. GOETHE

  • Prev by Date: Need help setting up a list of equations using "matrix notation"
  • Next by Date: Re: help number instruction
  • Previous by thread: Re:Alternative to Timing[10000!] benchmark
  • Next by thread: Re: Power Macs