Not many companies out there have as much simulatenous new projects going at the same time as Google seems to. The latest of their programs to be completed is called Caffeine. It promises to vastly increase the speed at which web indexing is done. (For those of you not in the know, web indexing is the process by which new websites are identified and cataloged for use by search engines.)
This whole Internet thing has really taken off, so with the greater expansion of information on it, the more new methods of classifying data have been needed. Google is fairly mum about the technology used behind the Caffeine process, but it seems like they are using a magnitude greater of parallel processors looking at smaller portions of the web, instead of larger and faster supercomputers progressing through entire fields of data in a more linear fashion. With Caffeine, Google's search index is updated "on a continuous basis", so that new sites will pop up much quicker now.
"Caffeine lets us index web pages on an enormous scale. In fact, every second Caffeine processes hundreds of thousands of pages in parallel. If this were a pile of paper it would grow three miles taller every second. Caffeine takes up nearly 100 million gigabytes of storage in one database and adds new information at a rate of hundreds of thousands of gigabytes per day. You would need 625,000 of the largest iPods to store that much information; if these were stacked end-to-end they would go for more than 40 miles," says the company.
As for the name -- well perhaps Caffiene can index Java-heavy websites? That's just a guess on our part.
If you'd like further information on how searching works, the video below discusses the basics of how Google indexes: