02-02-2017 08:13 AM - edited 02-02-2017 08:14 AM
I'm lately trying to optimize my LabVIEW FPGA code by reducing the minimum number of cycles that each while loops needs. The card I'm using is a RIO 7852R (40 MHz clock), and right now my bottleneck is the Analog I/O function, which apparently I cannot run inside a while loop faster than 100 cycles.
Is there any way to improve it? According to the specs it should be able run at 1 MHz, so around 40 cycles per loop.
Solved! Go to Solution.
02-02-2017 09:00 AM
Hello gerardpc,
can you provide a screenshot, how the benchmark was done or the FPGA code?
Is the analog in and out in the same loop?
This link could be helpful:
Optimizing your LabVIEW FPGA VIs: Parallel Execution and Pipelining
02-06-2017 07:55 AM
I found the problem. I was accessing the same Analog input from two different loops. This reduced the speed about two times.
After modifying my code with a FIFO instead of a dual direct access to my input I could reduce the number of clock cycles to about 50 (close to the theoretical optimum of 40).