Skip to content
Snippets Groups Projects
user avatar
Pietro Bonfa authored
6f4954e0
History
Name Last commit Last update
Tests
.gitignore
README.md
memory_report_new.f90
pw.x

New memory estimation for QE 6.1

The accuracy of the new routine for memory estimation (memory_report_new.f90) is evaluated on a series of test cases.

Test cases

The following pictures summarize the results that have been collected with gperftools. The absolute accuracy of the estimation seems to be better than 50MB/process with the notable exception of the simulation using real space algorithms for the charge augmentation where the estimation of the boxes points is very rough.

See below for a short comment on the various results.

Gamma point, Norm Cons., SCF, 1 pool (obviously)

FeNCGamma

This is an interesting case in which the initialization of the wavefunction is more memory demanding that the iterative solution. This is due to natomwfc being very large.

Gamma point, PAW, SCF, 1 pool

FePAWGamma

The peaks in this picture are from addusdens

Gamma point, PAW, SCF with tqr, 1 pool

FeNCGamma

In this case, the estimation of tab(...)%qr is off by almost 200 MB.

Gamma point, PAW, VC-RELAX, 1 pool

FeNCGamma

In this case addusforces sets the threshold for the required memory.

Gamma point, US, SCF, 1 pool

CNOHGamma

This test case is very small. The accuracy of the estimation is impacted by the small sized matrices that are not considered in the evaluation. Notably, MPI and MKL allocations becomes sizable in this case.

4 kpoints, US, SCF, nspin=2, 1 pool

Ni

Collinear magnetism, up to convergence.

4 kpoints, US, SCF, nspin=1, 1 pool

Ni

4 kpoints, US, SCF, nspin=2, 2 pool

Ni

4 kpoints, US, SCF, noncollin (nspin=4) , 2 pool

Ni