mergentia
machine intelligence beyond AI
SYNTHETIC INTELLIGENCE @ THE EDGE
THE CHALLENGE
Learning at the edge
-
90 ZettaBytes (>50% of all data) will be generated in edge IoT devices by 2025.
​​
-
ML and AI training at the edge are constrained by limited resources - compute, memory, and bandwidth. ​
​​
-
Compute demands for training in the cloud are unsustainable and double every 3-4 months.​
​
-
Data latency precludes real-time decision making, inflates costs and increases privacy and security risks.​​​
OUR SOLUTION
Synthetic intelligence
-
Offers a new paradigm of energy-efficient information processing in scalable and reconfigurable functional materials
​
-
Leverages self-organization in a brain-inspired, nanoscale architecture enabling real-time dynamic adaptation to edge data streams
​​
-
Pushes the limits of classical computation beyond current analog and neuromorphic hardware