The CAPO tool begins the process of parallelizing by importing serial code (at top left). The tool then analyses data dependencies to determine the relationship among variables in the code. After the data dependency analysis, a loop level analysis helps determine how to divide the code to run in parallel. Then, a series of graphical user interfaces guides users through the process of dealing with sections of code that cannot be parallelized in original form.