Fast, in-memory data stores already exists with support and ease of use/deployment...I was able to set up hazelcast for my distributed cluster in minutes...I don't see the advantage of this over nosql or in-memory data storage.