Parallel loops & improved efficiency 
Author Message
 Parallel loops & improved efficiency

The major concerns with parallel loops are just the added processing,
sharing data between parallel loops, and synchronization.  Race
conditions are often a concern when setting the value or properties of
variables in multiple parallel loops.  Most of these issues can be
avoided by careful programming and use of the synchronization tools
such as Queues and notifiers from the Advanced menu.  The processing
problem of having several parallel loops is of course minimized by
having sufficient delays in the loops.  In 6.1 it can be minimized
more by using the new event structure to eliminate the need for
polling front panel controls.

Maintaining a very large buffer of data is pricipally an issue of
usable RAM, but may also cause speed and/or graphics issues when
graphs need to be redrawn.  Beware of the fact that labview often
needs to replicate data behind the scenes, so you will often end up
using more memory than you thought you would.  Also, consider the
efficiency of how you are storing your data.  Remember that contiguous
blocks of memory are usually required for arrays, and therefore very
large arrays are more difficult to allocate.  You might consider using
some amount of file I/O to periodically store data and then retrieve
it as necessary.

Sun, 15 Aug 2004 00:09:47 GMT  
 [ 1 post ] 

 Relevant Pages 

1. Improve efficiency of 3 big array ops.

2. they improve efficiency?

3. question about efficiency of fortran in parallel environments

4. Efficiency and Nested Loops VS Readability

5. NumPy loop efficiency

6. Nested do loop efficiency question

7. Loop efficiency

8. Parallel Loops Question

9. parallel while-loops

10. parallel loop

11. progress && loop

12. Interpreters, efficiency, &c


Powered by phpBB® Forum Software