quantum-espresso/LAXlib
Pietro Bonfa 8457380d92 Aligned changes to touch-dummy removal. 2018-05-22 12:05:46 +02:00
..
Makefile removed ifdef from LAXlib makefile[?] 2018-03-06 16:09:48 +01:00
README.TEST - adding command line parameter "-n" to specify the matrix size 2016-02-13 19:30:39 +00:00
cdiaghg.f90 in parallel diagonalization eigenvalues are broadcast to the whole 2017-08-19 13:27:51 +00:00
distools.f90 Cleanup of unused routines. There are 1001 different routines distributing 2017-11-02 17:31:15 +00:00
dspev_drv.f90 ELPA cleanup, following a suggestion by Ye Luo, Argonne: define 2018-02-05 17:16:07 +01:00
la_error.f90 QE made ready to use MPI module instead of mpif.h include file. See the three 2016-11-25 14:11:28 +00:00
la_param.f90 QE made ready to use MPI module instead of mpif.h include file. See the three 2016-11-25 14:11:28 +00:00
la_types.f90 Cleanup of unused routines. There are 1001 different routines distributing 2017-11-02 17:31:15 +00:00
make.depend wrong checkout corrected for removing eigsolv form make.depend 2018-03-06 17:09:08 +01:00
mp_diag.f90 - interface with LAMMPS was broken in 6.2 2017-11-25 13:41:57 +00:00
ptoolkit.f90 QE made ready to use MPI module instead of mpif.h include file. See the three 2016-11-25 14:11:28 +00:00
rdiaghg.f90 in parallel diagonalization eigenvalues are broadcast to the whole 2017-08-19 13:27:51 +00:00
test.f90 __OPENMP => _OPENMP (set by all OpenMP-aware compilers) 2017-10-25 07:32:10 +00:00
transto.f90 QE made ready to use MPI module instead of mpif.h include file. See the three 2016-11-25 14:11:28 +00:00
zhpev_drv.f90 ELPA cleanup, following a suggestion by Ye Luo, Argonne: define 2018-02-05 17:16:07 +01:00

README.TEST

This library provides a Testing and Timing code
to asses the performance of Linear Algebra subroutines, 
estimate the scalability and the size of your simulation.

To compile the test program, once you have properly
configure QE, go inside the directory LAXlib and
type:
	make TEST

Then you can run your LA tests using command like:

	mpirun -np 4 ./la_test.x -n 1024

or, in case of serial build

	./la_test.x -n 1024

where 1024 is the size of the matrix to be processed.