i used to work for the company that now owns the music recognition algorithm that shazam is using. it's staggering how much hardware it takes to make this stuff

i used to work for the company that now owns the music recognition algorithm that shazam is using. it's staggering how much hardware it takes to make this stuff work. they have a database of millions of songs, all of which must be kept in ram at all times, to make recognitions fast. it requires hundreds of linux servers. despite my preference for stuff i've actually worked on, i think the midomi app works just as well or better. but they were late to the party, so they're having trouble getting the eyeballs.