Deep Learning Accelerators

Intigia is a design company specialized in IP cores and SoC design for safety critical applications
Our IP cores are written in human readable HDL code and vendor independent

Intigia’s engineering team develops vendor independent deep learning and machine learning accelerators to speed up the execution of Convolutional Neural Networks (CNN) and classical Computer Vision algorithms in small factors electronic boards.

Initigia research in Edge AI and TinyML is focused on developing hardware accelerators capable of real time processing with low power on SoCs. Intigia is also actively researching in functional safety applied to machine learning algorithms executed on FPGAs and SoCs.

Our designs have been successfully tested on field on drones and other applications.

Our portolio includes:

  • Convolutional Neural Network hardware accelerator
  • Recurrent Neural Network hardware accelerators
  • Hardawre accelerators for multiple OpenCV functions
  • Device drivers

Object detection and tracking for drones, automotive and surveillance.

Cloud image detection for space imaginary

Our IP cores are vendor independent. We have tested them on Xilinx devices Zynq 7 and Zynq Ultrascale.

Our development methodology is designed to minimize pitfalls and risks early in the project, based on automated quality assurance processes. Ask us about compliance with safety critical standards (DO-254, ISO 26262, ESCC).

Ask one of our experts how we can help you