When Artificial Intelligence meets next-generation Networks
Autonomous vehicles, drones ... The "new edges" acquire and process data, make decisions and connect to each other. The advent of 5G in mobile communications should amplify the phenomenon.
With the development of artificial intelligence, specific microprocessors (the GPUS or graphics processing units) allowed the emergence of machine learning or "deep learning". Originally designed for graphics cards and video gaming machines, they have well found a second life. Now they are found in data centers, as well as in autonomous vehicles, drones, robots, smart sensors. Each time, they interact with networks, which results in a new paradigm.
These "new edges" as Americans call them, acquire and process data, make decisions and also connect to other "edges". Autonomous vehicles, cars, trucks connect to each other, but also with drones for example. And the advent of 5G in mobile communications is expected to accelerate the movement by reducing latency and increasing throughput, thereby treating locally whatever will be. The power and storage in these new devices are such that they look like small data centers. Connections with the cloud still need to be updated to feed and teach the centralized system, which in turn, fuels the smart "new edge"devices thanks to increased speeds of 5G.
The "new edge" consumes power and speed is key. The amount of devices covered by this technology is already a bigger fleet than the sum of mobile phones (about 4.7 billion) and smartphones (about 2 billion). Analysts expect the range of 26 to 30 billion devices in 2020, and figure that the current base is between the 6-10 billion. The market for these networks 'new edge' that connect and manage boosted objects in artificial intelligence is a unique opportunity that will amount to several billion dollars.
Georges Nahon, CEO ORANGE SILICON VALLEY - Orange