11-27-2009 07:49 AM
Basically, I have got a 1024x1024 2d array (fixed size). What I do:
1. Perform an FFT along each line (1024 times)->this is fast
2. I split my big array in 8 sub-arrays to take advantage of my 8 cores
3. Rebuild my 1024x1024 2d array -> this is slow
To perform step 3 I used "Build Array" vi but also I firstly initialised and used the "Replace subarray" vi. Unfortunately it's still very slow. Could anyone suggest me a way to speed up the "rebuilding" of my 2d array, or do you have another solution to use all the cores without splitting up my big array?
Thanks a lot.
Solved! Go to Solution.
11-27-2009 07:53 AM
Hi Adrian,
use a FOR loop set for parallel processing.
Or mimic that automatism of LV by using the DataValueReference to work on the same memory in different (FOR) loops.
All this applies to LV2009 - you didn't note your LV version...
11-27-2009 08:44 AM
I do use LV2009, so I can set a FOR loop for parallel processing!
Now it's fast!
Thanks a lot!