I do not think Maple help on the subject is easy to follow as the exampels are not clear. I was hoping someone can give a very simple example.
Now, I run long script, over say 1,000 problems. This takes long time each time I run it. Since each problem is completely independent of others, and there is no shared data at all (for each problem, its output is written to separate file), I am thinking of using the parallel programming in Maple, and hoping this will speed it up. I have an intel PC, with modern CPU. I think it has may be 10 or 16 cores, not sure now, so in theory it should be faster to complete.
For a very basic example, suppose I have this sequential program
for i from 1 to 10 do
Where foo() is now called 10 times, one after the other. How would one change the above to make it run in parallel?
I know I need to call Threads:-Task:-Start and then use Threads:-Task:-Continue
But I am not sure how to use Threads:-Task:-Continue to tell it to call foo in parallel passing each task 1,2,3,.... in turn, and wait until they are all done. This is the part not clear to me how to do even after looking at the help example under Task Programming Model
Could someone show how to do the above for this simple example?
Since I have 1,000 problems, do I need to create 1,000 tasks at once, one for each problem and wait for them to all be done? what if I have 10,000 problems, creating 10,000 tasks at once might not work, would it? or will Maple handle this internally by queuing the creation of tasks as needed? This part is also not clear to me.