Weird memory profile from memory intensive FDTD simulations in MEEP!

[aa16oaslak]

Hey guys,

I am currently doing a memory intensive simulation using MEEP. I did a memory profiling of one of the sims and got a curve something like this:

My expectations was something similar to what's happening after time 11-11 09 (constant memory usage over time)-- Could you please let me know why I am seeing a curve like this?

I am using the mpirun command with 16 cores with split_chunks_evenly = False (so that resources will be allocated according to the computational costs) and eps_averaging = True

My simulation size is:

"cell_size": {
        "x": 1626 mm,
        "y": 830 mm,
        "z": 0},
"resolution": 9 pixels/mm

Looking forward to your reply!