How To Perform Synchronized Parallel Computing Using Parfeval?

1 view (last 30 days)
Hi ,
I am running 2 functions running with an infinite loop within using the parfeval function provided in the parallel computing toolbox. Is synchronization between these 2 functions possible? If yes, then is there any way to establish a semaphore mechanism or an external control dependency during the runtime of these functions ? I have been trying to get the output using fetchOutputs to return the signal the control which causes a blocking call and disrupts the execution of my code.
Is there any other way to achieve this goal using the parallel computing toolbox?
Thanks and regards,
Kaushik Gowda H N
  5 Comments
Mohammad Sami
Mohammad Sami on 24 Sep 2021
Edited: Walter Roberson on 25 Sep 2021
You cannot declare a global variable in main thread and expect its value to propagate between main thread and workers.
You can look at my answer below to see how you might setup a two way data exchange between main thread and a single worker. You may repeat that with another worker.
However I don't see how your code improves the efficiency, one of the worker will always be idle in your current design, since it depend on the other thread to calculate some value and change the state. In your design only one worker is active at one time.
You may want to see how the workers can work on independent calculations so that you can use both workers simultaneously.

Sign in to comment.

Answers (0)

Categories

Find more on Data Type Identification in Help Center and File Exchange

Products


Release

R2021a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!