This is an implementation of an answer to the following question:
A folder contains a large # (~50k) of files. Write a C# program that finds files with duplicate contents.
My answer was to do a first pass of all the files, take a md5 hash of a small chunk (4k bytes) of each file. Store 1st instance of hash & filename, all others with matching hash get their filename & the potential match filename added to a list of potential duplicates. On the second pass, do full MD5 hash of each potential duplicate file & compare. Output resulting match list.