...could use python's .split(' ') but then what do you plan to do with address ranges?
>>> addresses = ['1919 LAWSON RD','919 PRINCE CHARLES LN','1437-1455 PAYNE RD','407 N BRAINTREE DR']
>>> for address in addresses:
address.split(' ')
['1919', 'LAWSON', 'RD']
['919', 'PRINCE', 'CHARLES', 'LN']
['1437-1455', 'PAYNE', 'RD']
['407', 'N', 'BRAINTREE', 'DR']
>>>
>>> for address in addresses:
address.split(' ', 1)
['1919', 'LAWSON RD']
['919', 'PRINCE CHARLES LN']
['1437-1455', 'PAYNE RD']
['407', 'N BRAINTREE DR']
>>>
...so if that (above) is good enough for you, then it's a single statement:
arcpy.CalculateField_management("your fc or tbl", "your num add fld", '!your orig add fld!.split(" ")[0]', "PYTHON_9.3")
...and of course for the alternate field for the remaining text, run a 2nd execution (by the way, with an updatecursor you could run a single execution to write both fields)---
arcpy.CalculateField_management("your fc or tbl", "your txt add fld", '!your orig add fld!.split(" ")[1]', "PYTHON_9.3")
