GDL> time_compare, ['idl7ubuntu','g100913', 'g101107', 'g110914','g111011']
idl7ubuntu: |TIME_TEST3 performance for IDL 7.0:
idl7ubuntu: |       OS_FAMILY=unix, OS=linux, ARCH=x86
idl7ubuntu: |	Wed Oct 12 00:07:51 2011
   g100913: |TIME_TEST3 performance for IDL 6.0:
   g100913: |       OS_FAMILY=unix, OS=Linux, ARCH=i686
   g100913: |	Wed Oct 12 00:10:45 2011
   g101107: |TIME_TEST3 performance for IDL 6.0:
   g101107: |       OS_FAMILY=unix, OS=Linux, ARCH=i686
   g101107: |	Wed Oct 12 00:11:11 2011
   g110914: |TIME_TEST3 performance for IDL 6.0:
   g110914: |       OS_FAMILY=unix, OS=Linux, ARCH=i686
   g110914: |	Wed Oct 12 00:12:14 2011
   g111011: |TIME_TEST3 performance for IDL 6.0:
   g111011: |       OS_FAMILY=unix, OS=Linux, ARCH=i686
   g111011: |	Wed Oct 12 00:11:40 2011

    Time idl7ubu g100913 g101107 g110914 g111011
    0.03     100^    750*    768*    186*    161*  Empty For loop, 2000000 tim
    0.02     100^    627*    615*    267*    167*  Call empty procedure (1 par
    0.02     100^    624*    613*   1904*    155*  Add 200000 integer scalars 
    0.02     100^    562*    552*   1544*    230*  50000 scalar loops each of 
    0.01     100^   1831*   1846*    532*    294*  Mult 512 by 512 byte by con
    0.01     100^   3863*   3859*   4633*   4630*  Shift 512 by 512 byte and s
    0.02     100^   1882*   1958*    548*    308*  Add constant to 512x512 byt
    0.02     100^   2895*   2903*    659*    372*  Add two 512 by 512 byte arr
    0.02     100^    652*    650*    238*    115*  Mult 512 by 512 floating by
    0.05     100^    276*    274*    319*    315*  Shift 512 x 512 array, 60 t
    0.03     100^    826*    832*    210*    113   Add two 512 by 512 floating
    0.02     139*    123*    124*    100^    101   Generate 1000000 random num
    0.03     102     100     101     101     100^  Invert a 192^2 random matri
    0.00   12986*    120*   1300*    100^    100^  LU Decomposition of a 192^2
    0.02     100^   1419*   1372*    734*    273*  Transpose 384^2 byte, FOR l
    0.03     100^    563*    515*    349*    379*  Transpose 384^2 byte, row a
    0.05     100^    520*    521*    239*    261*  Transpose 384^2 byte, TRANS
    0.03     100^   1302*   1238*    796*    570*  Log of 100000 numbers, FOR 
    0.02     105     277*    279*    100^    101   Log of 100000 numbers, vect
    0.02     100^    215*    243*    172*    172*  131072 point forward plus i
    0.02     100^   3296*   4085*   2587*   2536*  Smooth 512 by 512 byte arra
    0.02     100^   1739*   1742*   1296*   1262*  Smooth 512 by 512 floating 
    0.02     100^    342*    443*    235*    111   Write and read 512 by 512 b
    0.56     100^    885*    911*    634*    442*  Total Time
    0.02     100^    530*    600*    328*    212*  Geometric mean
 
^ = fastest.
* = Slower by 15% or more.
Wed Oct 12 00:13:16 2011