Edge Questions

Most current analysis on Edge looks at promises and possibilities. I have data but no firm answers on key decisions. In particular:

Will the Edge naturally be inside the enterprise or inside the telco network or …?

Partial answer: Yes, it will be everywhere. But how will it divide?

Will the Edge on the telcos side be dominated by the carriers or the cloud giants, which now include Alibaba & Huawei?

Telcos and others are discovering that distributed clouds are hard. The webscale outfits employ thousands of engineers. Synchronization across hundreds of points in realtime is a challenge.

Who will buy?

One approach: what latency does a given app require? VR needs to be under 20 milliseconds, preferably closer to 10 ms. Most 5G networks are built to 30 ms, so they can’t support VR. Connected cars usually do fine at 50-100 ms, so don’t need 5G. Only when the server is driving the “autonomous” car are sub-10 ms required. The car companies aren‘t planning for that.

The size of the market won’t be clear until the networks are in full commercial service, not close in most of the world. Carriers are telling me that large volume facial recognition is the only application selling so far.

Google, Microsoft, and Amazon are deeply involved and looking to dominate. The telcos are looking hungrily at the possibility of revenue. Everyone’s experimenting with alliances without a clear path.

The complexity and cost will be challenging. These are hybrid networks, with truly challenging demands for timing and concurrency. The

All answers welcome.

Improvements welcome