Nikita Bedunkevich
Head of hardware development OneSoil

Minsk, Belarus
Being a student, Nikita started developing children's educational applications, and then changed his line of work to do what he loved. He was engaged in the development of automation for grain drying complexes, and later on the heating control system for private houses. In 2016, he began working at OneSoil and participated in its creation. He is engaged in the development of data transmission systems between on-board computers of agricultural machinery and a system for monitoring soil and weather parameters on the farmers' fields. A year ago, the long-time enthusiasm for FPGA technology was developed into a project to create an accelerator for rendering neural networks on chips with a small capacity.

Report: Machine learning in IoT devices.
Overview of existing solutions and architectures for embedded systems. FPGA-based neuro-accelerator implementation for IoT.
  • How to apply ML in IoT now?
  • How do systolic arrays work in FPGA? And why has Google chosen them?
  • Why do neural networks in embedded systems not work as well as on GPU?


© All Right Reserved. My company Inc.
e-mail us: