Hi,
My lmr simulations seem to take too much memory to run.
SLURM's seff command reports that a steady simulation with 14.2 M cells, ideal gas model, and SA turbulence model used about 1950 GB.
The same but unsteady simulation with "euler" gasdynamic update scheme used about 1860 GB of memory.
In contrast, a steady SU2 simulation with 80 M cells, ideal gas model, and SA turbulence model used about 830 GB.
Why is there such a huge difference in the required memory between Eilmer and SU2? Why did the Eilmer unsteady sim take so much memory if it had to store variables only at 2 time steps? I am wondering if there is something I can change in the attached lua file to significantly reduce the memory usage.
cube.txt
Hi,
My lmr simulations seem to take too much memory to run.
SLURM's seff command reports that a steady simulation with 14.2 M cells, ideal gas model, and SA turbulence model used about 1950 GB.
The same but unsteady simulation with "euler" gasdynamic update scheme used about 1860 GB of memory.
In contrast, a steady SU2 simulation with 80 M cells, ideal gas model, and SA turbulence model used about 830 GB.
Why is there such a huge difference in the required memory between Eilmer and SU2? Why did the Eilmer unsteady sim take so much memory if it had to store variables only at 2 time steps? I am wondering if there is something I can change in the attached lua file to significantly reduce the memory usage.
cube.txt