Dear Mohamed,
TSP function which is very helpful function to speedup your calculations
Note that strictly speaking, the “spawn” feature is what allows you to speedup your calculation: you create jobs (“spawns”), which can possibly run in parallel.
The task spooler (“tsp”) instead, is simply in charge of the distribution of those jobs among the cores available on your machine.
If you stay on “default” (no “tsp”), jobs will still run in parallel on a multi-core machine (leading to speedup). However, you will have no control on how the jobs are distributed.
Now, how does tsp distribute jobs on your machine? It simply relies on a maximum number of jobs N which can run simultaneously. It creates N queues, among which the jobs are dispatched.
Of course, N <= cores in your machine.
I used through the terminal the command tsp -S 4. Then in FLAIR I select “tsp”
With tsp -S 16 in the command line, you set the maximum number of jobs which can run simultaneously to 16.
When you select “tsp” with Flair, this upper limit will be used (here, 16).
what does tsp04, … or tspN refer to?
Alternatively, you can set that maximum number of jobs which can run simultaneously directly inside of Flair.
“tsp04” sets an upper limit of 4 simultaneous jobs.
“tsp08” sets an upper limit of 8 simultaneous jobs.
“tspN” sets the number of cores available in your machine as an upper limit.
how I recognise that Flair uses this function
In a terminal, “tsp -l” should display the list of jobs, along with their status (queued / running / finished).
Hope this helps,
Cheers,
Gabrielle
