DHFR benchmark of Amber 7 on Macintosh
According the method came with the amber package, I did two benchmarks, this time it would be more serious than previous one and also comparable to Amber 7 Benckmarks page.
The measurement should be more accurate because the time step is longer. Please check the table below.
Special thank to professor Eric Mockensturm at Penn State University for his help on G5 machine.============================================================================= "jac" == Joint Amber/Charrm DHFR benchmark. This is the protein DHFR, solvated with TIP3 water, in a periodic box. There are 23,558 total atoms, and PME used with a direct space cutoff of 9 Ang. This is the benchmark in benchmarks/jac subdirectory of the Amber 7 distribution. -------------------------------------------------------------------------------- node-name CPU OS compiler npcu time-per-step -------------------------------------------------------------------------------- mjhsieh 400 MHZ G4 MacOS X g77 3.4 1 5.644 Prof. E. M. 2 GHZ G5 (dual) MacOS X g77 3.4 1 1.275 Prof. E. M. 2 GHZ G5 (dual) MacOS X XLF 8.1b 1 0.777 ============================================================================= "gb_mb" == Generalized Born myoglobin simulation. This protein has 2492 atoms, and is run with a 20 Ang. cutoff and a salt concentration of 0.2 M, with nrespa=4 (long range forces computed every 4 steps.) This is the test case in the benchmarks/gb_mb subdirectory of the Amber 7 distribution. -------------------------------------------------------------------------------- node-name CPU OS compiler npcu time-per-step -------------------------------------------------------------------------------- mjhsieh 400 MHZ G4 MacOS X g77 3.4 1 4.857 Prof. E. M. 2 GHZ G5 (dual) MacOS X g77 3.4 1 0.697 Prof. E. M. 2 GHZ G5 (dual) MacOS X XLF 8.1b 1 0.543
Labels:
Compiler
