Blade Server Racks

When would you use a blade server, rack mount server and a standalone server?

Which type of companies would use each type and why.
Also if you have have any other information on each type, as i have to talk about each one in a presentation in collage 2morro :L

Then for a talk, I suggest that you start looking in Wikipedia etc

as a 30 second answer here, will not keep you going long..
but in general each is optimised for what it does 'not' nead.

Blades : have computing power but almost no memory storage (mem or and no disk), and use remote disk arrays.. They remotely boot. access disk/disk arrays (often in RAID - go look that up also), for multi/array processing, reliable applications.

Rack mount: simplifies equipment rooms, but has disk, has memory..
but very little room for I/O (liek a blade) so gigabit ethernet or simple serial/Scsi IO.. but completelky standalone for small applications eg DNS servers, Data in/data out with no big need for databases or temprary diskspace.

Standlone : has room for lots more disks....more I/O space/options/card slots and direct connectivity but needs space.. and cannot usually be easily stacked (or are already pre-cabineted).. and even when they can...they take up space.
But multi processor ones, can have very high speed and redundant bus structures internally for failover.

its all about what can you live without...