MathGroup Archive 1995

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

Search the Archive

Speed foe Mma ans Win95

  • To: mathgroup at smc.vnet.net
  • Subject: [mg2366] Speed foe Mma ans Win95
  • From: jpk at apex.mpe.FTA-Berlin.de (Jens-Peer Kuska)
  • Date: Sun, 29 Oct 1995 22:47:53 -0500

Hi, Windows 95 & Mathematica users,

test results for Mathematica 2.2 and Windows 95

System: 486 DX100 (AMD) 16 MByte Ram
-----------------------+--------+----------+-----------
                       | Win95  | Win95 &  | HIGH pri
                       |        | AddTwo   | to kernel
-----------------------+--------+----------+-----------
 Timing[               |        |          |
   N[                  |        |          |
     Sin[1/2],         |        |          |
     2500              |        |          |
    ]                  |        |          |
  ][[1]]               | 28.559 | 23.564   | 11.492
-----------------------+--------+----------+----------
 Timing[               |        |          |
   N[                  |        |          |
     Pi,               |        |          |
     10001             |        |          |
    ]                  |        |          |
   ][[1]]              | 50.90  | 35.770   | 22.977
-----------------------+--------+----------+----------
 Timing[10001!][[1]]   | 37.154 | 35.287   | 19.264
-----------------------+--------+----------+----------
 First[                |        |          |
   Timing[             |        |          |
     Eigenvalues[      |        |          |
       Table[          |        |          |
         Random[],     |        |          |
         {200},        |        |          |
         {200}]]]]     | 68.852 | 54.110   | 31.717
-----------------------+--------+----------+----------
f[x_]:= 4x-4x^2;       |        |          |
Timing[                |        |          |
  Nest[f,0.6,5000]     |        |          |
 ][[1]]                | 23.971 | 18.571   | 10.176
-----------------------+--------+----------+----------
f[x_]:= BesselJ[0,x];  |        |          |
Timing[                |        |          |
  Nest[f,0.6,2500]     |        |          |
 ][[1]]                | 51.729 | 39.788   | 21.394
-----------------------+--------+----------+----------
Timing[                |        |          |
  Simplify[            |        |          |
    kdv[q3]            |        |          |
  ]                    |        |          |
 ]                     |187.101 | 157.063  | 88.778
-----------------------+--------+----------+----------

The process priority of the kernel
was increased manualy with the ps.exe / nice.exe pair 
form the unix95.zip archiv. 

One can find it on

ftp.cica.indiana.edu:/pub/pc/win32/sysutil/unix95.zip

The other tasks and the system was still usable, especially
the frontend works fine.

I had make only one run with every configuration, always
with a fresh loaded kernel.

All othe tasks had normal priority.

Jens

 


  • Prev by Date: using Math Link w/FORTRAN
  • Next by Date: Re: wierd REmove behavior
  • Previous by thread: using Math Link w/FORTRAN
  • Next by thread: Paolo's question