We attempt to optimize the computer time we use. In particular, we prepare numbers of TD-runs using Octopus software.

Here we keep a small note on the performances reached on various computers.

Example of Silicon primitive cell

MachineOctopus VersionCompiler & ParametersTime per SCF stepTime per TD step
CPUGPUCPUGPU
IBM Power9 (1 node + 1xGPU 16 GB)Octopus 10.4 (distrib. binary)GCC/GFortran1539 s500 s (single-kgrid, StatePack=no) 1860 s 115 s (StatePack=no)
IBM Power9 (1 node + 4xGPU 16 GB)Octopus 10.4 (distrib. binary)GCC/GFortran 36 s
(single-kgrid, StatePack=no)
 –22 s (single-kgrid, StatePack=no)
IBM Power9 (1 node + 4xGPU 16 GB)Octopus 10.4 (distrib. binary)GCC/GFortran  113 s
(single-kgrid, StatePack=yes)
 –22 s
(single-kgrid, StatePack=yes)
IBM Power9 (1 node + 4xGPU 16 GB)Octopus 10.4 (distrib. binary)GCC/GFortran 139 s
(4x kgrid, StatePack=no)
xx s
(4x kgrid, StatePack=no)
 MPCDF Draco (4 nodes) Octopus (version of 2019 09 11) Intel compilers  – 2.5 s –
IT4I Salomon (4 nodes)Octopus (version of 2019 06 10)Intel compilers 5.0 s
IT4I Barbora (8 nodes)Octopus (version of 2019 06 10)Intel compilers 1.91 sxx s
 PRACE Prometheus (Poland) (4 nodes) Octopus (version of 2019 10 16) Intel compilers  5.0 s –

Example of silica primitive cell

  • dx=0.22 Bohr. k=8³. RAM: 22 GB. Duration per SCF cycle on Barbora: <72 s.