Nick Destrycker (Edgise)

Building custom AI hardware for edge computing

11:50-12:30

Artificial intelligent algorithms are increasingly being implemented on-device instead of in the cloud. Edgise is a venture that develops custom hardware for edge computing/edge AI to efficiently run complex AI models. In this talk, we’ll go over a practical use case, advantages, disadvantages and why edge computing is preferred in some circumstances. We’ll give an understanding of what’s inside an edge device, how it works and which hardware is appropriate to build and run AI algorithms. We’ll give a broad introduction to the current edge AI landscape. Afterward, we’ll dig deeper and explain the what and how of enhancing edge devices by accelerating compute-intensive algorithms in hardware.

Throughout his electronic engineering studies, Nick Destrycker nurtured his passion for digital chip design. In the meantime, his interest in artificial intelligence grew. After graduation, he worked as a tape-out engineer at Imec and as a digital design engineer at ICsense. In his pursuit to combine hardware development with AI, he co-founded Edgise.

Nick Destrycker

Edgise