EC4600
AI Edge Computer
EC4600 AI edge computer for real-time visual recognition and local image analytics.
PHOTO
AI Edge Computer for Real-Time Visual Recognition
Edge AI compute
14 TOPS, octa-core ARM, Adreno GPU
Vision-ready I/O
4× camera power, HDMI, USB, SPK
Flexible connectivity
4G, Wi-Fi 6, 6× FE, dual SIM
Industrial design
Metal shell, IP40, fanless, wide
temp
Stable operation
HW/SW watchdog, link detect, BAT
-
Core Technical Specifications
AI Computing Power 14 TOPS (comprehensive); octa-core ARM up to 2.7 GHzCellular Network LTE Cat4, 4G; 2 × Nano SIM; dual SIM failoverDimensions (W × D × H) 190 × 143 × 40 mm (including mounting parts); wall mountingEthernet Ports 6 × 10/100 MbpsNetwork Features ARP, Ethernet; Static IP, DHCPOperating System Android 12Operating Temperature & Protection -10 °C ~ +60 °C; IP40; fanless metal housingPower Supply 12 V DCReliability Dual SIM backup; multi-level link detection; embedded watchdogSecurity Secure BootVision & Serial I/O 4 × camera power (5 W); 2 × RS-232; 2 × RS-485; HDMI; 2 × USB 2.0Wi-Fi STA; 802.11a/b/g/n/ac/ax; 2.4 / 5 GHz dual band
-
Hardware Performance
AI Computing Power 14 TOPS (comprehensive)CPU Qualcomm QCM6490: 1 × Kryo 670 Gold Plus @ 2.7 GHz + 3 × Kryo 670 Gold @ 2.4 GHz + 4 × Kryo 670 Silver @ 1.9 GHzFlash 128 GB eMMCGPU Qualcomm® Adreno™ 643 @ 812 MHzRAM 8 GB
-
Hardware Interface
Backup Battery 1 × backup battery interfaceCamera Power Supply 4 × white round interface (Ø5.5 × 2.1), 5 WEthernet Port 6 × 10/100 MbpsGPIO 2 × OUTHDMI 1 × HDMINTC 1 × NTC temperature sensingPower Button 1 × PowerSerial Port 2 × RS-232 (DB9); 2 × RS-485 (one 4-pin with power supply, one 3-pin industrial terminal)SIM 2 × Nano SIM slotSpeaker 1 × SPK, supports 5 W speakersUSB 2 × USB 2.0
-
Wi-Fi
Frequency Band 2.4 / 5 GHz dual bandOperating Mode STAProtocol 802.11a/b/g/n/ac/ax
-
LED Indicators
3G/4G Cellular indicator always on when connected normallyPower Power supply indicator, always on after power-onStatus Status indicator flashes during normal operationUser LED 4 × LED
-
Power
Power input 12 V DC
-
Mechanical
Cooling FanlessDimensions (W × D × H) 190 × 143 × 40 mm (including installation parts)Housing MetalIngress Protection IP40Mounting Wall mounting
-
Environmental
Ambient Humidity 5 ~ 95 % RH (non-condensing)Operating Temperature -10 °C ~ +60 °CStorage Temperature -40 °C ~ +85 °C
-
Physical
EFT Level 2ESD Level 2Free Fall IEC 60068-2-32Shock IEC 60068-2-27Surge Level 2Vibration IEC 60068-2-6
-
Operating System
OS Android 12
-
Network Features
LAN Protocol ARP, EthernetNetwork Type 4GWAN Protocol Static IP, DHCP
-
Security
Secure Boot Supported
-
Reliability
Backup Dual SIMEmbedded Watchdog Device self-diagnosing; auto-recovers from operation faultsLink Detection Multi-level link detection; auto redials once disconnected
This product is currently unavailable.
Standard package*
- EC4600 AI Edge Computer *1
- 4G Antenna *2
- Wi-Fi Antenna *2
- GNSS Antenna *1
* Standard package contents vary by standard order codes.
Optional accessories
- Power Supply *1
Contact us directly to learn more about our packaging options.
DeviceLive
IoT Device Management Platform
Device management, remote monitoring, edge app management, and remote maintenance to help industrial enterprises build intelligent edge networks.
Frequently Asked Questions
What does the EC4600 offer?
The EC4600 is a Qualcomm-based AI edge computer purpose-built for real-time visual recognition in industrial and commercial deployments. Highlights:
| Compute | Qualcomm QCM6490 octa-core ARM Cortex (1× Kryo670 Goldplus @ 2.7 GHz + 3× Gold @ 2.4 GHz + 4× Silver @ 1.9 GHz); Qualcomm Adreno 643 @ 812 MHz GPU; 14 TOPS comprehensive AI compute |
| Memory | 8 GB RAM, 128 GB eMMC |
| OS | Android 12 |
| Cellular | 4G LTE Cat 4; 2× nano-SIM with dual-SIM failover |
| Wi-Fi | 802.11 a/b/g/n/ac/ax (Wi-Fi 6), 2.4 GHz + 5 GHz dual-band, STA mode |
| Ethernet | 6× 10/100 Mbps |
| Camera I/O | 4× camera-power supply outputs (5521 white round connectors, 5 W each) for IP-camera deployments |
| Other I/O | 2× RS-232 (DB9), 2× RS-485 (one 4-pin with power, one 3-pin terminal), 2× USB 2.0, 1× HDMI, GPIO, NTC temperature input |
| Audio / status | 1× SPK (drives 5 W speaker), 4× LEDs, backup-battery interface, power button |
| Power / form factor | 12 V DC; metal housing, fanless, IP40, wall-mount, 190 × 143 × 40 mm |
| Environment | −10 to +60 °C operating |
- EC4600-DQ20-STD — China LTE Cat 4 (LTE FDD B1/3/5/8, LTE TDD B34/B38–41, WCDMA B1/B8, TD-SCDMA B34/B39, CDMA BC0, GSM 900/1800)
- EC4600-FQ58-STD — EMEA LTE Cat 4 (LTE FDD B1/3/7/8/20/28A, LTE TDD B38/40/41, WCDMA B1/8, GSM B3/8)
How do I install and access the EC4600?
- Wall-mount the EC4600 using the integrated mounting points (190 × 143 × 40 mm metal chassis, IP40, fanless).
- Connect antennas for 4G (SMA) and Wi-Fi (RP-SMA), insert SIM card(s) into the dual nano-SIM slots.
- Connect cameras — up to four IP cameras can draw power from the 4× 5521 white-round camera-power outputs (12 V, 5 W per output) and stream over the EC4600’s Ethernet ports.
- Apply 12 V DC power. The Power LED is solid red on power-up; the Status LED flashes during normal operation; the 3G/4G LED is solid when cellular is connected.
- Access Android 12 via HDMI display + USB peripherals, or remotely via SSH once the device is on the network.
How do I wire and power IP cameras to the EC4600?
The EC4600 exposes two distinct interfaces for IP-camera deployments:
| Connection | Source on EC4600 | Spec |
| Camera power | 4× camera-power outputs | 5521 white-round connector, 5 W per output |
| Camera data | 6× 10/100 Mbps Ethernet ports | Standard RJ45 |
Up to 4 cameras can be powered directly from the camera-power outputs; the Ethernet ports carry the camera streams, with the remaining ports available for the wired uplink or other devices. On Android 12, configure your AI-vision application to consume the camera streams over the local network.
For runtime resilience, the EC4600 includes independent hardware and software watchdogs that auto-recover from operation faults, plus a backup-battery interface that keeps the device running through power interruptions.
How do I deploy AI visual recognition on the EC4600?
The EC4600’s 14 TOPS of comprehensive AI compute (across the Kryo670 CPU and Adreno 643 GPU) runs visual-recognition models locally on Android 12, eliminating cloud round-trips for real-time decisions. Typical pipeline:
- Connect up to 4 IP cameras to the camera-power outputs (5 W each at 5521) and the Ethernet ports.
- Deploy your recognition app as an Android APK or AI-runtime container on Android 12.
- Process locally — the SoC’s ML acceleration runs your visual-recognition algorithm on the live camera streams.
- Push results to a third-party cloud management platform via the cellular / Wi-Fi / wired uplink for device-status monitoring, fault alerts, and online status analysis.