ATMA comes with a nifty DupeDetector module which lets you load a bunch of files and compare the items in them for dupes. You may dump the duped item info to a text file. The format of this file is looks like the regular dump info file. The File Selector for the DupeDetector has the MULTISELECT option turned on so you can use 'shift' and 'ctrl' to select multiple files at once.
Note: The DupeDetector now loads the files incrementally and runs each one through the detector as needed instead of pre-loading everything, so it should be faster and better behaved than previous versions. You can now give it the option of loading the data without graphics. This means that as soon as it is done with the data, it releases the memory associated thereby making it less of a memory hog. This is now the preferred way of doing things.