What makes google search results within milliseconds

[someonehateblue]

Please help me. Is there someone knowing inside process of google when they do searching? (any aspect will be very useful)
Thank you

[troels nybo nielsen]

Welcome to WW, someonehateblue.

The point is that Google (and other search engines as well) have a very large database on their computers. And a lot of fast computers to find the search results in that database. Many of those results may be outdated, but Google's freshbot updates search results on many webpages every day. That is on web pages that are known to be updated frequently.

[dormek]

Linux, Linux and Linux :) There are 20,000 Redhat servers...

If you know Spanish:
[google.dirson.com...]

HTHY

[vitaplease]

[webmasterworld.com...]

an old sound-track, but worth the listening..

[Markus]

The most important thing should be that almost everything is preprocessed. Check this document to see what they did in the early days to provide search results efficiently:

[dbpubs.stanford.edu:8090...]

The key is the inverse index and the above paper shows some ideas about how to sort document IDs in the index. All this stuff, combined with some 10,000s of servers, makes the difference.

[Robert Charlton]

I've also heard in a radio interview that they use RAM instead of hard drives, so results are served up a lot faster.