Top Page > Browsing

about "Automatic running test with large-scale systems" (openmx3.9) Date: 2021/03/30 17:34 Name: Hiromi Hayashi   <hiromi@hpc-sol.co.jp> I am running a large-scale test with openmx3.9. ==runtest.result 1 large_example/5_5_13COb2.dat Elapsed time(s)= 46.53 diff Utot= 0.000000000020 diff Force= 0.000000000009 2 large_example/B2C62_Band.dat Elapsed time(s)= 539.83 diff Utot= 0.000000000120 diff Force= 0.000000011798 3 large_example/CG15c-DC-LNO.dat Elapsed time(s)= 76.13 diff Utot= 0.000000000271 diff Force= 0.000000000280 4 large_example/DIA512-1.dat Elapsed time(s)= 49.14 diff Utot= 0.000000002723 diff Force= 0.000090520226 5 large_example/FeBCC.dat Elapsed time(s)= 223.33 diff Utot= 0.000000016680 diff Force= 0.000000077938 6 large_example/GEL.dat Elapsed time(s)= 55.04 diff Utot= 0.000000000012 diff Force= 0.000000000006 7 large_example/GFRAG.dat Elapsed time(s)= 32.40 diff Utot= 0.000000000004 diff Force= 0.000000000003 8 large_example/GGFF.dat Elapsed time(s)= 694.50 diff Utot= 0.000000001038 diff Force= 0.000000000172 9 large_example/MCCN.dat Elapsed time(s)= 75.25 diff Utot= 0.000000000042 diff Force= 0.000000000186 10 large_example/Mn12_148_F.dat Elapsed time(s)= 82.28 diff Utot= 0.000000000098 diff Force= 0.000000000009 11 large_example/N1C999.dat Elapsed time(s)= 1197.62 diff Utot= 0.000000000993 diff Force= 0.000000267336 12 large_example/Ni63-O64.dat Elapsed time(s)= 69.30 diff Utot= 0.000000063888 diff Force= 0.000001291813 13 large_example/Pt63.dat Elapsed time(s)= 101.82 diff Utot= 0.000000187429 diff Force= 0.000000046646 14 large_example/SialicAcid.dat Elapsed time(s)= 27.53 diff Utot= 0.000000000002 diff Force= 0.000000000002 15 large_example/ZrB2_2x2.dat Elapsed time(s)= 199.10 diff Utot= 0.000000000010 diff Force= 0.000000000030 16 large_example/nsV4Bz5.dat Elapsed time(s)= 272.39 diff Utot= 0.000000505073 diff Force= 0.000000178336 The difference result in 44 and 12 is large. I don't know what the problem is. The options I made are as follows. All have almost the same results. Intel＋MKL+IntelMPI CC = mpiicc -fopenmp -O3 -march=native -I${MKLROOT}/include/fftw FC = mpif90 -fopenmp -O3 -march=native -I${MKLROOT}/include/fftw LIB= -L${MKLROOT}/lib/intel64 -lmpifort -lmkl_scalapack_lp64 -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -lmkl_blacs_intelmpi_lp64 -liomp5 -lpthread -lifcore -lm -ldl -lgfortran aocc+MKL+openmpi CC = mpicc -I${MKLROOT}/include/fftw -O3 -march=znver2 -mtune=znver2 -mfma -mavx2 -m3dnow -fomit-frame-pointer -fopenmp FC = mpif90 -I${MKLROOT}/include/fftw -O3 -march=znver2 -mtune=znver2 -mfma -mavx2 -m3dnow -fomit-frame-pointer -fopenmp LIB = -L${MKLROOT}/lib/intel64 -lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64 -lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core -lgfortran -lpthread gcc+MKL+IntelMPI CC = mpicc -O3 -fopenmp -ffast-math -I${MKLROOT}/include/fftw FC = mpif90 -O3 -fopenmp -ffast-math -I${MKLROOT}/include/fft LIB = -L${MKLROOT}/lib/intel64 -lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64 -lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core -lgfortran -lpthread Any advice would be appreciated. Hiromi.

Page: [1]