[strlen]

Does the result of this research end up as products?

Microsoft Research is excellent as a research organization, but it seems that it's goal isn't to develop new products based on this research but rather to make sure that the scientists aren't developing products based on their research elsewhere.

Note: there's also lots of research work that needs to be done that isn't going to be applied immediately. That's certainly fine -- but there's been research that has went on at Sun that has been turned into industry changing products (from NFS to Java).

[sern]

Yes. Off the top of my head, IBM invented hard disks, relational databases and RISC. They definitely aren't as innovative now as they were thirty years ago, but at least they have more to show than Oracle does.

[strlen]

From what I've read, IBM was a top-notch R&D company up until the late 90s/early 2000s (with all the innovations you mentioned, as well as literally writing the book on software development -- they were in no way a pure research shop). Since then they've transformed into a professional services company, however (although they've always been known for having great sales/professional services).