[MITgcm-support] Question about cpl_aim+ocn running too slowly

[闫明羽]

Dear All,




I successfully run the coupled model (/verification/cpl_aim+ocn), but it takes more time to finish the simulation (about 5 times slower than Marshall et al. 2007 and Brunetti et al. 2019).




The command I have used to run the coupled experiment is "mpirun -np 1 ./build_cpl/mitgcmuv : -np 1 ./build_ocn/mitgcmuv : -np 1 ./build_atm/mitgcmuv". 




In the instruction file (/verification/cpl_aim+ocn/run_cpl_test), it seems that only 1 core can be used to run the CPL, and if I add the number of cores to run the atmosphere or ocean, it will be slower.  




Do you have any suggestions to improve the speed of model running?




Thank you in advance,




Mingyu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20210314/17fb5a29/attachment.html>