Network requirements for latency-critical services in a full cloud deployment

The end-point of the cloud trend is that all computational and storage resources leave the homes and move to the cloud. We can be rather certain the future eventually looks like this as IT and telecommunications services are commoditised and the end users' time, interest and skill to acquire and maintain electronics fade away. The network architecture suitable for such scenario will be analysed in terms of latency critical services to establish delay requirements and feasible localisations of the data centers. It is concluded that for cloud gaming as an example of a time-critical service, the maximum delay is limited to 20 ms. In the network architecture this corresponds to locating the data center in the Main CO, i.e. typically in the aggregation part of a metropolitan network.