Because the IMAQ-Vision function to display an image was slow for large images I wrote my own routine using the Windows SDK and it was much faster.
However, now that the application has been moved to a dual processor PC the opposite is true, i.e. the application runs much faster if I use the IMAQ-Vision function. Inspection reveals that this is due to equal loading on the two processors, whereas if I use my SDK routine all the load is on the first processor.
Is there some explanation for this? Do the IMAQ-Vision routines have built in multi-threading?