|
|
From James:
|
|
|
|
|
|
-----------
|
|
|
|
|
|
I have measured the amount of time taken for each set of interactions between cells: corner, edge, face and the amount of time taken for the self interactions.
|
|
|
|
|
|
I have performed a fixed dt run for 128 time steps:
|
|
|
|
|
|
`./swift_fixdt -t 16 -f CosmoVolume/cosmoVolume.hdf5 -m 0.6 -w 5000 -c 1e-5 -d 1e-7 -e 1e-7`
|
|
|
|
|
|
which produced the following statistics:
|
|
|
|
|
|
`
|
|
|
Corner Interaction: 1.68%
|
|
|
Edge Interaction: 10.56%
|
|
|
Face Interaction: 40.15%
|
|
|
Self Interaction: 47.61%
|
|
|
`
|
|
|
|
|
|
I have also run a multiple time step simulation for 28000 time steps:
|
|
|
|
|
|
`./swift -t 16 -f CosmoVolume/cosmoVolume.hdf5 -m 0.6 -w 5000 -c 1. -d 1e-7 -e 0.01`
|
|
|
|
|
|
which produced the following statistics:
|
|
|
|
|
|
`
|
|
|
Corner Interaction: 12.43%
|
|
|
Edge Interaction: 27.58%
|
|
|
Face Interaction: 30.12%
|
|
|
Self Interaction: 29.87%
|
|
|
`
|
|
|
-------------
|
|
|
Raw data here:
|
|
|
[fixdt_interaction_timings.dat](/uploads/aa50b317866667ea9821af642550393f/fixdt_interaction_timings.dat) |
|
|
\ No newline at end of file |