next up previous contents index
Next: Input file Up: User's manual of OpenMX Previous: Automatic running test   Contents   Index

Automatic running test with large-scale systems

In some cases, one may want to know machine performance for more time consuming calculations. For this purpose, an automatic running test with relatively large-scale systems can be performed by

For the MPI parallel running

     % mpirun -np 128 openmx -runtestL
For the OpenMP/MPI parallel running
     % mpirun -np 128 openmx -runtestL -nt 2
Then, OpenMX will run with 16 test files, and compare calculated results with the reference results which are stored in 'work/large_example'. The comparison (absolute difference in the total energy and force) is stored in a file 'runtestL.result' in the directory 'work'. The reference results were calculated using 16 MPI processes of a 2.6 GHz Xeon cluster machine. If the difference is within last seven digits, we may consider that the installation is successful. As an example, 'runtestL.result' generated by the automatic running test is shown below:

1 large_example/5_5_13COb2.dat Elapsed time(s)= 39.43 diff Utot= 0.000000000013 diff Force= 0.000000000046
2 large_example/B2C62_Band.dat Elapsed time(s)= 572.22 diff Utot= 0.000000000025 diff Force= 0.000000013928
3 large_example/CG15c-Kry.dat Elapsed time(s)= 40.71 diff Utot= 0.000000002112 diff Force= 0.000000001090
4 large_example/DIA512-1.dat Elapsed time(s)= 37.93 diff Utot= 0.000000169524 diff Force= 0.000000033761
5 large_example/FeBCC.dat Elapsed time(s)= 81.55 diff Utot= 0.000000000649 diff Force= 0.000000001349
6 large_example/GEL.dat Elapsed time(s)= 47.05 diff Utot= 0.000000000066 diff Force= 0.000000000002
7 large_example/GFRAG.dat Elapsed time(s)= 24.05 diff Utot= 0.000000000122 diff Force= 0.000000000015
8 large_example/GGFF.dat Elapsed time(s)= 639.31 diff Utot= 0.000000000051 diff Force= 0.000000000243
9 large_example/MCCN.dat Elapsed time(s)= 53.72 diff Utot= 0.000000009994 diff Force= 0.000000016474
10 large_example/Mn12_148_F.dat Elapsed time(s)= 76.58 diff Utot= 0.000000000096 diff Force= 0.000000000090
11 large_example/N1C999.dat Elapsed time(s)= 97.56 diff Utot= 0.000000006902 diff Force= 0.000000007356
12 large_example/Ni63-O64.dat Elapsed time(s)= 78.00 diff Utot= 0.000000000782 diff Force= 0.000000000047
13 large_example/Pt63.dat Elapsed time(s)= 60.40 diff Utot= 0.000000002147 diff Force= 0.000000000059
14 large_example/SialicAcid.dat Elapsed time(s)= 47.80 diff Utot= 0.000000000005 diff Force= 0.000000000003
15 large_example/ZrB2_2x2.dat Elapsed time(s)= 143.16 diff Utot= 0.000000000030 diff Force= 0.000000000003
16 large_example/nsV4Bz5.dat Elapsed time(s)= 104.20 diff Utot= 0.000000010770 diff Force= 0.000000000605
Total elapsed time (s) 2143.68

The comparison was made using 128 MPI processes and 4 OpenMP threads (totally 256 cores) on CRAY-XC30. Since the automatic running test requires large memory, you may encounter a segmentation fault in case that a small number of cores are used. Also the above example implies that the total elapsed time is about 36 minutes even using 256 cores. See also the Section 'Large-scale calculation' for another large-scale benchmark calculation.

next up previous contents index
Next: Input file Up: User's manual of OpenMX Previous: Automatic running test   Contents   Index
t-ozaki 2013-05-22