Dear developers,
I just run the test-suite to see if my QE is ok.
The cp passed all test.
but the pw met some problem.
The cp passed all test.
but the pw met some problem.
- for serial execution
I got 177 out of 181 tests passed (2 unknown). and I notice that error in one fail is significant.
pw_pawatom - paw-vcbfgs.in: **FAILED**.
Different sets of data extracted from benchmark and test.
More data in benchmark than in test: ef1.
p1
ERROR: absolute error 2.20e-01 greater than 1.00e-01. (Test: -0.52. Benchmark: -0.3.)
e1
ERROR: absolute error 4.00e-05 greater than 1.00e-06. (Test: -328.23191. Benchmark: -328.23187.)
n1
ERROR: absolute error 2.00e+00 greater than 1.50e+00. (Test: 4.0. Benchmark: 6.0.)
especially -0.52 vs -0.3 and 4.0 vs 6.0, is such significant error OK? - for parallel execution
it is much worse. I got only 127 out of 181 tests passed (2 unknown)
there are also some significant fails, like
pw_atom - atom-pbe.in: **FAILED**.
n1
ERROR: absolute error 2.00e+00 greater than 1.50e+00. (Test: 5.0. Benchmark: 7.0.)
pw_atom - atom-sigmapbe.in: **FAILED**.
n1
ERROR: absolute error 4.00e+00 greater than 1.50e+00. (Test: 15.0. Benchmark: 11.0.)
and many more. The full version see here https://pastebin.com/3TYFjebG
again, I am wondering are these significant error ok?
I compiled QE6.1 with intel parallel studio 2017u2, and use following configuration
mklopt="-mkl=sequential" && flagopt="-O3 -xHost -traceback -g" && ./configure dir=none MPIF90=mpiifort --with-scalapack=intel SCALAPACK_LIBS="-lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64" BLAS_LIBS="$mklopt" LAPACK_LIBS="$mklopt" FFT_LIBS="$mklopt" FFLAGS="$flagopt" CFLAGS="$flagopt"
best regards
_______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
