However its performance at the edge remains constrained by power, memory, and compute intensity. The future of AI lives at the edge, but it is currently being held back by legacy hardware.

There are four main challenges hindering modern day AI at the edge
We work to make it real!

The memory intensity of deep neural networks

The energy consumption of deep neural networks which drain the battery life of the device it is running on

Latency in inference because of the large size of deep neural networks

Latency from having to rely on the cloud to compensate for the computational limitation of edge device