At EdgeCortix, we automate the design of AI accelerators for on-device deep learning. Our Dynamic Neural Accelerator IP, and performance optimized AI Software & Hardware Co-exploration framework, reduces time to market and brings massive cost savings across multiple industry verticals.

Learn More

Our automated co-exploration engine uses AI to design AI chips, bringing cloud-level performance to resource constrained devices, for low latency, low cost and energy-efficient deep neural network inference.



EdgeCortix's DNA processor architecture and industry first  AI hardware and software co-exploration engine enable adaptation of AI accelerator designs with focus on edge devices to operate in real-time, at significantly reduced cost, power consumption and size.

Our software stack seamlessly integrates with standard deep learning development environments, with the ability to optimize models once and deploy flexibly across existing platforms, including our AI specialized accelerator, Intel x86 architecture, Nvidia GPUs and Arm CPUs. Bringing a complete solution from efficient deep learning architecture search to edge deployment across different market segments.

Ask About IP Licensing or Demo

Perception for Autonomous Vehicles, Advanced Driver Assistance Systems

Public safety, Traffic monitoring & surveillance, Video analytics

Predictive maintenance, inspection, machine vision

Intelligent assistance, voice recognition, battery optimization

Intelligent sensing & navigation, infrastructure inspection

Object tracking, security & surveillance, responsive retail advertising

Real-time energy efficient inference with DNA


Sakya Dasgupta, PhD

Founder & CEO

Founded in 2019 in Singapore and Japan, EdgeCortix is on a mission to rethink chip design using AI driven automation, specifically geared towards multiple orders of magnitude improvement in energy-efficient, low cost, small time to market, edge devices with real-time deep learning capability.

We have the ambitious goal of enabling a distributed intelligent edge ecosystem driven by our efficient chip design, that complements a centralized cloud infrastructure. All this while keeping data private, enabling real-time compute and finally the ability to learn continuously from the real-world data. Using a novel
co-exploration framework and adaptive AI accelerator design, we achieve this by bringing innovative solutions to enterprises, and make edge devices intelligent while optimizing for memory, compute and energy while seamlessly integrating with existing AI infrastructure
and frameworks.


Daniel K. Fujii

President & CEO
Trust Capital Co. Ltd
Blackstone Japan (past)

Norishige Morimoto

CTO & Vice President R&D
IBM Japan

Yiyu Shi, PhD

Director Sustainable Computing Lab
University of Notre Dame, USA



Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.