Im using a cluster and trying to monitor my gpu ussage. the command I'm using is:
srun --jobid 'jobid' watch -n 1 nvidia-smi
This then creates an output that looks like this: Broken Borders
As you can see the borders are all messed up.
I then ran the same code just wihtout the watch part:
srun --jobid 'jobid' nvidia-smi
and then the out put was good and looked like this: Normal Borders
I doesnt really bother me all that much since I can still tell how much gpu I am using, but I'm just a bit curious as to why this is happening.