Dear all,
I have checked a similar question brought on Oct 19, 2020. But still, I have a few questions. I would be very grateful if you could offer some help.
- When I runed
#!/bin/bash
#SBATCH -J raxml-ng-mpi
#SBATCH -N 1
#SBATCH -c 70
#SBATCH -p normal
#SBATCH --output=%x-%j.out
raxml-ng-mpi -all --msa Lac_core_alignment_2.fasta --model LG+I+G+F --bs-trees 500 --threads 70
, I received a CPU core oversubscription error (see the attached raxml-ng-mpi-327098.out). My questions are: 1) How can I know if it's a false alarm? 2) Since the result of RAxML recommended 68 threads to be used (see the attached T1.raxml.log), why I can't follow the recommendation and run the analysis?
- So I added '--force perf_threads' as:
#!/bin/bash
#SBATCH -J raxml-ng-mpi
#SBATCH -N 1
#SBATCH -c 70
#SBATCH -p normal
#SBATCH --output=%x-%j.out
raxml-ng-mpi -all --msa Lac_core_alignment_2.fasta --model LG+I+G+F --bs-trees 500 --threads 70 --force perf_threads
. Please see the attached output file raxml-ng-mpi-327100.out. My questions are: 1) Do you think it's appropriate to simply add `--force perf_threads` to run it? 2) Will you suggest I to `play around with thread pinning` as the last similar question?
Looking forward to any suggestions.
Best,
Nanzhen