YACC: Yet Another Crypto Conference
Fast Software Encryption (FSE) is a conference devoted exclusively for, as the name suggests, fast encryption in software. It was formed primarily because, at the time (early 90s), the Crypto and Eurocrypt conferences were not too interested in new algorithms (they still aren't, on the whole), and it was felt that having a conference devoted exclusively to this topic would produce some interesting new designs. There is a very strange dynamic between the different conferences, in that for the most part, new algorithm designs are almost looked down upon. This is in many ways understandable - we have dozens of really good block ciphers, so why create a new one? As to why Eurocrypt and Crypto are so heavily into theory, I honestly don't know. It's possible that it just makes for better research, which I can actually believe - anyone, from a novice up to an expert, can create a cipher that they themselves cannot break. But breaking someone else's stuff requires intelligence, talent, and a lot of hard work.
Perhaps the single most well known algorithm produced by FSE is Schneier's Blowfish algorithm, which has become very widely used around the world. He proposed Blowfish as a starting point for developing an encryption algorithm to replace DES; this process did occur, and eventually resulted in the AES competition (which the block cipher Rijndael won, beating out 14 competitors, including Schneier's Twofish).
Another intesting FSE algorithm is SEAL, designed by Rogoway and Coppersmith. It is an exceptionally clever design, and quite fast in software. However, due to a patent on SEAL (held by IBM), it has seen little use, which is quite a shame.
Literally dozens of other algorithms have been proposed at various FSE conferences over the years. Most of them have been broken. A few see wide use. FSE has also helped produce several new techniques of cryptanalysis, which would often be invented explicitly to attack an algorithm proposed at FSE.
FSE is held yearly somewhere around the world (the first was at Cambridge), often in conjunction with another conference (such as the 2000 AES conference in New York City).