OK people, I've created an signal processing algorithm that uses an input signal (typically 50,000 or so points) and outputs a small 2D array with the information I'm looking for.
It runs fairly well on my Host PC (Intel P4 at 3.2 GHz running LV 8.2 on Windows XP) with a computation time of about 20ms. However, when I implement the same algorithm on my RT system (using the exact same code), I get somewhere around 85ms. The RT system is a PXI-8187 (2.5 GHz P4-M). Is the system architecture THAT much better on my desktop PC?
There is no DAQ going on, just some FFT and then a bunch of array operations and some loops... standard stuff. I'm sure that this algorithm can be sped up some, but I'm wondering why my RT system is 4x slower doing the same calculations.
I've attached the VI of my algorithm. In general, what causes significant slow-down on the RT system?