Could you tell a bit more about your task besides the size of the file? Do you want to process all or most of the data in the xml file in some way, such as making aggregate statistics or converting to a different format? Or do you instead want to find only a few nodes that are easy to recognize in the XML without too much extra processing?