Search partition pruning in distributed databases is one concrete application I'm aware of.

Hashtables optimized for really large item counts; it can be a useful fast pre-check for item existence.