Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

TASK GEOMETRY/EFFICIENCY on Bluevista

I need to run CCSM3 on bluevista as efficiently as possible. I built the code using default values in the setup/run scripts as suggested in the user manual but it appears the code is 42% slower. I would like to keep the configuration as close to standard as possible.

The steps in the tutorial produce

#BSUB -n 104 and
setenv LSB_PJL_TASK_GEOMETRY "{(0,1,2,3,4,5,6,7)(8,9,10,11,12,13,14,15)(16,17,18,19,20,21,22,23)(24,25,26,27)(28,29,30,31,32,33,34,35)(36,37,38,39,40,41,42,43)
(44,45,46,47,48,49,50,51)(52,53,54,55,56,57,58,59)(60,61,62,63,64,65,66,67)
(68,69)(70,71)(72,73)(74,75)}"

My limited understanding is that this implies that I am asking for 104 processors and 13-nodes. What I am wondering is whether the last 4 parenthesis (68,69)(70,71)(72,73)(74,75) mean that these 4 nodes only have two tasks assigned; how can that be when bluevista has 8 processors on each node?

Second the suggestion for increasing throughput/reducing costs described here (http://www.cisl.ucar.edu/docs/bluevista/ccsm.html) refer to ccsm3_1_beta39 tags yet the latest code at ESG is beta14.

Finally where can I get information to help interpret the information printed when you run getTiming.csh so I can know when I have got improvements on the task geometry?
 
Top