Overview
The EC3320 series AI edge computer is designed for industrial AI inference and edge computing applications. Powered by the high-performance Rockchip RK3588 octa-core processor with an integrated 6 TOPS NPU, it delivers powerful local AI computing for machine vision, predictive maintenance, and smart surveillance. With rich interfaces including triple Gigabit Ethernet, dual HDMI, 4x USB 3.0, and industrial I/O, it seamlessly connects to cameras, PLCs, and sensors. Built-in Linux system and optional Hailo-8 AI module (up to 26 TOPS) provide flexible AI acceleration. InHand DeviceSupervisor™ Agent service enables easy data acquisition and cloud connectivity, while DeviceLive cloud management simplifies batch deployment and maintenance.
Powerful AI Edge Computing with 6 TOPS NPU
The EC3320 features the Rockchip RK3588 octa-core processor (4x Cortex-A76 + 4x Cortex-A55 up to 2.4GHz) with a built-in 6 TOPS NPU supporting INT4/INT8/INT16 mixed operations. Compatible with TensorFlow, MXNet, PyTorch, and Caffe frameworks, it delivers efficient AI inference for real-time machine vision, defect detection, and facial recognition. Optional Hailo-8 M.2 AI module extends computing power up to 26 TOPS for demanding multi-camera analytics workloads.
Rich Interfaces for Diverse Industrial Scenarios
The EC3320 offers a comprehensive interface portfolio: 3x 10/100/1000Mbps Gigabit Ethernet ports for multi-camera PoE connectivity, 2x RS-232 and 2x RS-485 serial ports for PLC communication, 4x USB 3.0 for high-speed camera and peripheral connectivity, 2x HDMI 2.0 for dual-display output, 4DI+4DO for industrial control, 1x CAN 2.0 for vehicle/automotive applications, plus MIC and SPK interfaces for audio applications. The M.2 expansion slot supports both SATA3.0 SSD storage and Hailo-8 AI module upgrades.
InHand Device Supervisor™ Agent Service
The EC3320 features the built-in InHand Device Supervisor™ Agent service, simplifying configuration and eliminating the need for programming when uploading data to the cloud. It seamlessly supports 80+ mainstream industrial protocols including Modbus RTU/TCP/ASCII, OPC UA, EtherNet/IP, ISO on TCP, Mitsubishi MC, FINS UDP, and Siemens PPI. Additionally, it offers compatibility with major public cloud platforms such as AWS, Azure, and Alibaba, as well as seamless integration with mainstream IoT cloud platforms, providing users with diverse IoT cloud solutions.
Fully Industrial-Grade Design for Harsh Environments
The device is engineered with a highly reliable design specifically for harsh industrial environments. It meets EMC Level 3 standards and operates efficiently across a wide temperature range from -20 to 60°C, while supporting a wide voltage power supply of 9 to 36VDC. The fanless metal housing with IP40 protection ensures durability in industrial settings. Independent hardware and software watchdog mechanisms guarantee automatic recovery from operation faults, ensuring worry-free 24/7 operation.
Uninterrupted Internet Access with 5G and Dual SIM
The EC3320 offers multiple network access options for reliable data transmission: It supports global 5G NR (SA/NSA) and 4G LTE Cat6 cellular networks, drawing on InHand's 20 years of experience in cellular network communication to ensure high-availability internet access, even in locations without wired networks. Additionally, it supports triple Gigabit Ethernet and Wi-Fi 5 connectivity, with the ability to mutually back up wired, cellular, and Wi-Fi networks, along with support for dual SIM failover. This redundant design of WAN links ensures uninterrupted and reliable internet access for mission-critical AI applications.
InHand DeviceLive Cloud Services Streamline Batch Deployment and Maintenance
They offer cloud-based parameter configuration, container management, edge AI application management, and firmware management, simplifying the setup and maintenance process. Cloud-based status monitoring allows users to easily check online status, signal strength, traffic consumption, and NPU utilization at a glance. Additionally, they provide safe and reliable remote access services for PLCs and other user equipment, enabling remote monitoring and management with peace of mind. Remote batch management of edge AI applications enables seamless model updates across the entire fleet.
Product Compare
| Edge Computer |
|
|
 |
| Model | EC5350 | EC942 | EC3320 |
| CPU | 6-core Cortex-A78AE @ 1.5GHz | Quad-core Cortex-A55@2.0GHz | Octa-core (4xA76 + 4xA55) @ 2.4GHz |
| GPU | 1024-core NVIDIA Ampere | None | Mali-G610 MC4 |
| NPU | 40 TOPS | None | 6 TOPS, extendable to 26 TOPS (Hailo-8) |
| RAM | 8GB | 4GB | 8GB LPDDR4X |
| FLASH | 128GB SSD | 16GB eMMC | 64GB eMMC |
| EtherNet Port | 2*10/100/1000Mbps | 2*10/100/1000Mbps | 3*10/100/1000Mbps |
| Serial Port | 2*RS232/485/422, DB9 | 2*RS-232/485/422, DB9 | 2*RS232 + 2*RS485, DB9 |
| CAN | 1*CAN FD | CAN 2.0A/B | 1*CAN 2.0 |
| SIM Card | 2*Nano SIM | 2*MicroSIM | 2*Nano SIM |
| Dimensions | - | 47.3*162.7*148.3mm | 180x136x54mm |
Certifications
