AI roadmap: the future of edge AI

Artificial intelligence (AI) is no longer the sole domain of sci-fi enthusiasts or computer scientists. It’s generally acknowledged as a force that will fundamentally change all aspects of our lives. We’re seeing an explosive growth in the adoption of AI to address a wide range of problems.

But the story has only just begun. So far, most AI implementations are examples of cloud AI. Here, extremely powerful computers crunch enormous quantities of data. In the future, however, the quest for intelligence will also be pursued at the edge.

Need for innovation

At the edge, trillions of devices will work together to create inconceivable amounts of new data. Most of these ‘things’ will be extremely constrained when it comes to power budget, cost, robustness, ...

If we want these devices to meet the performance demands of AI within these constraints, we will need innovation on all levels: from computational concepts down to the implementation technology.

This white paper sketches the roadmap for these edge AI innovations. Dive into it for a visionary view of the path towards next-generation AI applications.

Download now

What you will learn from this AI roadmap

Read this white paper to find out:

• Why it makes sense to make a separate roadmap for AI at the edge.
• What are the current computational models and architectural choices for AI: from (Bayesian) machine learning to neuromorphic and neuro-inspired computing.
• Which metrics could be used to objectively compare technology options.
• What are the technological opportunities for the future: 3D integration and interconnect, new memory and logic concepts, ...
• Which application moonshots can drive development: from human avatars to swarms of autonomous drones