I am impressed with the relationship managed in graph data modeling. I from SQL and NoSQL backgroud, neo4j provides the simple clustering mechanism, hope it supports ACID, CAP at a time.
I am digging deep into it, on how it solves the problems of RDB...
