http://qs321.pair.com?node_id=11118393


in reply to Could SHA encode faster than MD5?

If the only thing that you need to know is that "the data is probably different," and you expect that if it is different it will be very different, then a simple CRC-32 or CRC-64 digest is often enough. It is extremely fast. See preshing.com/20110504/hash-collision-probabilities/