Here is the benchmark script . It calculates 4x10^8 whole chains with the number of independent calculations set to 100 for a 156nt DNA with a Jfactor of ~2500 nM. You can compare your resulting output files to these to insure the calculations are working correctly. Remember, because of the Monte Carlo (random number generated) nature of the calculation, your results may vary slightly from these. It is best to run the benchmark a few times (the script runs it twice) and average the results.
On the chart below, to explain some of the more obtuse columns titles: 'CPU mfg' = CPU manufacturer, 'Comp' = compiler, 'Opts' = optimizations used in the compile, 'OS' = operating system, 'Gener' = benchmark from generate_chains, 'Analy' = benchmark time from analyze_chains, and 'Name' = the common name of the computer.
CPU mfg CPU model MHz Comp Opts OS Gener Analy Total(s) Name ------- --------- ---- ----- ---- -------- ----- ----- ----- ------- Intel Celeron 550 pgcc 2 lnx 2.2 maul Intel Celeron 550 gcc 1 lnx 2.2 109 314 423 maul Intel PII 400 pgcc 2 lnx 2.0 145 287 432 corona Intel PII 400 gcc 1 lnx 2.0 155 363 518 corona Intel PPro 200 pgcc 2 lnx 2.0 290 500 790 bass Intel PPro 200 gcc 1 lnx 2.0 324 645 969 bass SGI R5000 200 cc 3 irix 6.3 336 820 1156 ernst SGI R4400 150 cc 3 irix 5.3 636 1482 2118 tecate SGI R4000 100 cc 3 irix 5.3 960 2322 3283 dosequis
Opt 1: '-O3 -fast-math'
Opt 2: '-fast'
Opt 3: '-O2 -mips2'
Compilers: