Matrix Multiplication

Back to the Win32 Shootout Back to dada's perl lab

The test parameter N is the number of times we multiply 2 matrices.

Measurement of CPU as N varies   N Test Source 1 100 200 300 vc 0.01 0.03 0.04 0.07 vc++ 0.01 0.03 0.05 0.07 delphi 0.01 0.03 0.05 0.07 mingw32 0.01 0.03 0.06 0.09 gcc 0.03 0.05 0.07 0.10 ocaml 0.01 0.04 0.08 0.12 bcc 0.01 0.04 0.09 0.12 slang 0.04 0.07 0.10 0.14 lcc 0.01 0.05 0.10 0.15 vpascal 0.01 0.07 0.12 0.19 gnat 0.01 0.07 0.14 0.20 gforth 0.04 0.10 0.17 0.23 fpascal 0.01 0.08 0.16 0.23 pike 0.16 0.19 0.23 0.26 csharp 0.10 0.21 0.34 0.45 java 0.45 0.54 0.64 0.73 nice 0.52 0.63 0.74 0.84 se 0.01 0.38 0.76 1.14 pliant 0.35 0.79 1.24 1.73 smlnj 0.04 0.76 1.46 2.18 poplisp 0.09 1.94 3.82 5.69 ocamlb 0.04 2.41 4.81 7.20 erlang 0.50 2.82 5.16 7.50 lua5 0.04 2.82 5.64 8.45 lua 0.05 4.13 8.23 12.35 oz 0.20 4.26 8.35 12.47 python 0.26 6.02 11.84 17.67 ici 0.09 6.54 13.05 19.59 tcl 0.12 7.72 15.40 23.09 perl 0.16 11.48 22.90 34.31 guile 0.25 11.56 22.95 34.37 cygperl 0.17 11.52 22.97 34.41 php 0.25 17.89 35.62 53.27 awka 0.23 18.28 36.45 54.57 jscript 0.35 18.44 36.69 54.94 ruby 0.34 25.50 51.10 76.90 gawk 0.36 30.28 60.51 90.64 mawk 0.40 35.87 71.64 107.42 rexx 1.56 151.23 T T ghc 0.01 F F F vbscript 0.36 T T T

Measurement of Memory as N varies   N Test Source 1 100 200 300 vc 464 464 464 464 vc++ 512 512 512 512 delphi 600 600 600 600 mingw32 516 516 516 516 gcc 1520 1520 1520 1520 ocaml 680 680 680 680 bcc 576 576 576 576 slang 1844 1848 1848 1848 lcc 516 516 516 516 vpascal 576 576 576 576 gnat 748 748 748 748 gforth 1508 1508 1508 1508 fpascal 536 536 536 536 pike 5408 5408 5408 5408 csharp 3424 3428 3424 3424 java 4596 4596 4596 4596 nice 4928 4932 4932 4932 se 576 576 576 576 pliant 3224 3448 3880 4252 smlnj 936 936 936 936 poplisp 3300 3300 3300 3300 ocamlb 380 380 380 380 erlang 5280 5264 5264 5264 lua5 660 776 776 776 lua 708 1108 1108 1108 oz 652 652 652 652 python 2424 2436 2436 2436 ici 1320 1552 1552 1552 tcl 1568 1596 1596 1596 perl 1572 1604 1604 1604 guile 2484 2488 2488 2488 cygperl 2472 2496 2496 2496 php 3164 3280 3284 3280 awka 1884 1884 1884 1884 jscript 2528 2528 2528 2528 ruby 2044 2936 2936 2936 gawk 2280 2280 2280 2280 mawk 1744 1744 1744 1744 rexx 1976 66908 T T ghc 1144 F F F vbscript 2360 T T T

The following graph is similar to the graph on the index page, but here, N is much smaller, and startup costs may be more prominent.

[Note: Values have been normalized to fall in the range of 0-10 for aesthetic reasons. Original value ranges are included on the X-axis.

Source Code CPU (sec) Mem (KB) Lines Code Log delphi 0.01 600 40 log mingw32 0.01 516 52 log lcc 0.01 516 52 log vpascal 0.01 576 57 log ghc 0.01 1144 27 log se 0.01 576 70 log vc 0.01 464 51 log vc++ 0.01 512 52 log fpascal 0.01 536 59 log gnat 0.01 748 0 log bcc 0.01 576 51 log ocaml 0.01 680 37 log gcc 0.03 1520 52 log lua5 0.04 660 37 log ocamlb 0.04 380 37 log gforth 0.04 1508 21 log smlnj 0.04 936 80 log slang 0.04 1844 17 log lua 0.05 708 35 log poplisp 0.09 3300 44 log ici 0.09 1320 35 log csharp 0.10 3424 41 log tcl 0.12 1568 46 log pike 0.16 5408 27 log perl 0.16 1572 34 log cygperl 0.17 2472 41 log oz 0.20 652 66 log awka 0.23 1884 33 log guile 0.25 2484 38 log php 0.25 3164 34 log python 0.26 2424 34 log ruby 0.34 2044 37 log pliant 0.35 3224 42 log jscript 0.35 2528 40 log gawk 0.36 2280 33 log vbscript 0.36 2360 37 log mawk 0.40 1744 33 log java 0.45 4596 43 log erlang 0.50 5280 42 log nice 0.52 4928 0 log rexx 1.56 1976 38 log