NASE/HiL – Neural Architecture Search Engine with Hardware in the Loop

The data produced at the edge brings many opportunities for novel AI applications. Since the volume of produced data is increasing every year, we observe a shift away from the cloud and towards the edge for AI processing. But implementing AI on the edge is challenging, since computational resources are limited.

NASE/HiL finds and optimizes AI models for the edge automatically. It uses genetic evolutionary algorithms together with hardware awareness to optimize AI for almost any hardware platform. NASE/HiL boosts design phase productivity, reducing the development time from month to weeks compared to regular design processes. It is a revolutionary new way to work with AI, that increases performance, reduces energy consumption and saves development time for your models!