[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: matlab/octave performance hit
From: |
Steve Thompson |
Subject: |
Re: matlab/octave performance hit |
Date: |
Thu, 18 Oct 2007 15:26:16 -0700 |
As a data point ... my CentOS 4.5 laptop (1.7 GHz processor, 2 GB
RAM) running both
Matlab, Version 7.4.0.336 (R2007a), and
GNU Octave, version 2.1.57 (i686-pc-linux-gnu),
results in:
N = 200
Matlab = 0.000064
Octave = 0.023521
0.023521 / 0.000064 = 367.52
N = 1e6
Matlab = 0.075629
Octave = 0.32375
0.32375 / 0.075629 = 4.2807
N = 2e7
Matlab = 1.667574
Octave = 4.0107
4.0107 / 1.667574 = 2.4051
when computing `tic; test = 1 : N; cumsum (test); toc'.
So, as David guessed, the overhead is much less dramatic as N
gets bigger.
Steve
RE: matlab/octave performance hit, Bhaskar Mehta, 2007/10/18