Note the results in the following table result from a single
run of fcpu_time_test.pro: I haven't
tried to get the most accurate numbers by averaging over multiple
runs. Also, the number of bits as reported may, or may not reflect
the capabilities of the underlying CPU. In fact it is the value
stored in IDL's "!version.MEMORY_BITS" variable and may be
set by the version of IDL (32 vs 64 bits). As of 20090203, all
results reflect results obtained in Jan 2009 with the
“fcpu_time_test, fact=20” command.
Machine |
CPU |
Speed (GHz) |
N_CPU |
N_core/CPU |
Total Time (sec) |
Geometric Mean of 23 (sec) |
Comments |
columba (2/2009) |
Intel(R) Core(TM) i7 CPU 940; 64 bit |
2.93 |
8 |
4 |
9.383 |
0.310 |
IDL 6.3 |
eppob (1/2009) |
Intel/T5670 Duo; 32 bit |
1.80 |
1 |
2 |
19.87 |
0.726 |
IDL 7.0 |
cedar (1/2009) |
Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz; 32 bit |
2.40 |
4 |
4 |
13.66 |
0.476 |
IDL 6.3 |
earth (1/2009) |
Intel(R) Pentium(R) D CPU 3.20GHz; 32 bit |
3.20 |
4 |
2 |
22.28 |
0.851 |
IDL 6.3 |
gaia (1/2009) |
Intel(R) Pentium(R) D CPU 3.20GHz; 32 bit |
3.20 |
4 |
2 |
21.82 |
0.844 |
IDL 6.3 |
nstar (1/2009) |
Intel(R) Core(TM)2 Quad CPU Q9450 @ 2.66GHz; 32 bit |
2.66 |
4 |
4 |
12.46 |
0.449 |
IDL 6.3 |
peto |
Intel(R) Pentium(R) 4 CPU 2.20GHz; 32 bit |
2.20 |
1 |
1 |
42.12 |
1.655 |
IDL 6.3 |
chara |
Intel(R) Pentium(R) 4 CPU 3.00GHz; 32 bit |
3.00 |
1 |
1 |
33.00 |
1.300 |
Idl 6.3 |
firefly |
Intel(R) Pentium(R) D CPU 3.40GHz; 32 bit |
3.40 |
1 |
2 |
20.46 |
0.792 |
IDL 6.3 |
abmuloc(1/2009) |
AMD/ Opteron 252; @ 2.5 GHz 64 bit |
2.52 |
2 |
1 |
25.33 |
0.961 |
IDL 7.0 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|