Si Hammond
@stelleg are you running on Bowman?
@bathmatt - the size is the Kokkos allocations, high water is the max so far of that, the high-water-processor is the maximum resident set size (RSS) as defined by Linux...
Sounds like you have many small allocations? We can look at that.
This will not be fixed by SST v6.0.
@calewis - yes, I think we need to work out how to checkpoint long running on-node (CPU, memory, accelerator) simulations but low priority. Ideally, simulations are fast enough we never...
@ErsinCK - can you look in the config.log file which is generated when `./configure` runs (it will be in the same directory). There should be some operations to check for...
@ErsinCK - the `config.log` correctly states that `-lmpi` cannot be found. This will be the case if you do not have an MPI variant installed (e.g. MPICH, MVAPICH, OpenMPI etc)....
Hi @tommasobo - how do you normally run MPI on your cluster? SST automatically detects the number of MPI ranks being given to it to use and will distribute the...
OK, I am using the 2.0.3 release. I changed the output code at the end and it corrects the FOM (but note - I didn't change the number of elements...
I can submit a pull request as I get these fixed.