How to improve computational efficiency in Abaqus Cae

I am building a big model to select some specific elements using Python script in Abaqus Cae. The script runs quickly and the result is as expected in a small model. However, the computational efficiency in a bigger model is very low. The script has run 18 days in a Linux workstation with noGUI command. Are there any methods to set parallel computation in Abaqus cae? Thanks in advance.