Edgehax provides industrial-grade compute modules and single-board computers that integrate powerful processing, networking, and storage for edge AI applications. These solutions enable OEMs and innovators to accelerate the development and deployment of physical AI experiences for robotics, industrial gateways, and IoT devices.
Funding
Funding not disclosed
Founders
Product
Problem
Developing physical AI experiences for robotics, industrial gateways, and IoT devices requires integrating complex compute, networking, and storage capabilities onto compact hardware. This integration process is often time-consuming and resource-intensive for Original Equipment Manufacturers (OEMs) and innovators.
Solution
Edgehax provides industrial-grade compute modules and single-board computers (SBCs) designed for edge inferencing applications. These integrated solutions combine powerful processing units (CPU, GPU, NPU) with robust networking options (4G/LTE, Wi-Fi, LoRa) and high-speed storage (DDR with UFS). This all-in-one approach enables OEMs and innovators to rapidly develop and deploy physical AI experiences, reducing development cycles and accelerating time-to-market for products like humanoid robots, industrial gateways, and consumer IoT devices. The platform supports a range of AI workloads, from 2 TOPS to 140 TOPS (INT8), and offers features for remote deployment and monitoring.
Target Audience
The primary customers are OEMs, startups, system integrators, and innovators looking to build physical AI applications for robotics, industrial automation, automotive, drones, and consumer IoT devices.
Features
- Integrated compute (CPU, GPU, NPU) for edge AI workloads, offering performance from 2 TOPS to 140 TOPS (INT8).
- Multi-network connectivity including 4G/LTE, Wi-Fi, and LoRaWAN for flexible deployment and communication.
- Onboard DDR with UFS for accelerated data storage, model training, and inference.
- Support for multiple camera inputs, including up to 16 MIPI-CSI or 4 USB cameras, for vision AI applications.
- SDK provided for training and running Convolutional Neural Networks (CNNs) and transformer models for edge Generative AI applications.
- Features for remote deployment, Over-The-Air (FOTA) updates, and remote monitoring capabilities.
- Industrial-grade SBCs designed for rugged environments and diverse use cases such as robotics, Industry 4.0, automotive, and consumer IoT.
- Optional integrated NavIC GPS for centimeter-level navigational accuracy in vehicle and drone applications.