Documentation
TutorialsCommunity
  1. ALPON
  2. ALPON X5 AI

Overview

ALPON X5 AI · Overview · Updated 2026-05-06

The industrial edge AI computer.

ALPON X5 AI pairs a 25 TOPS DEEPX DX-M1 NPU with a Raspberry Pi Compute Module 5 host, multi-network connectivity, hardware root of trust, and centralized fleet management through ALPON Cloud. Built for real-time, on-device inference in production.

CES 2026 Best of Innovation 25 TOPS NPU Raspberry Pi CM5 LTE Cat 4 + eSIM 2× Gigabit Ethernet TPM 2.0 -20 °C to +60 °C
Quickstart Buy ALPON X5 AI
Sixfab ALPON X5 AI industrial edge AI computer with cellular and Wi-Fi antennas mounted on a rugged aluminum enclosure
What is the ALPON X5 AI?

ALPON X5 AI is Sixfab's industrial edge AI computer for production deployments. It runs on-device inference on a 25 TOPS DEEPX DX-M1 NPU, alongside a Raspberry Pi CM5 host (quad-core Cortex-A76 at 2.4 GHz), NVMe SSD, and always-online connectivity (2× Gigabit Ethernet, Wi-Fi, LTE Cat 4 with eSIM, Bluetooth 5.0). The fanless aluminum enclosure operates from -20 °C to +60 °C at IP40 and is managed centrally through ALPON Cloud. Intelligented by DEEPX. Built on Raspberry Pi.

Intelligented by DEEPX Built on Raspberry Pi CES 2026 Best of Innovation

At a glance

The numbers that define ALPON X5 AI as a production edge AI platform.

AI Performance
25 TOPS DEEPX DX-M1 NPU · INT8
Host CPU
2.4 GHz Quad-core Cortex-A76 · CM5
Memory
8 / 16 GB LPDDR4 ECC · NVMe SSD
Connectivity
4 Network Paths 2× GbE · LTE Cat 4 · Wi-Fi
Power input
9–30 V DC USB-C PD 27 W · PoE+ optional
Security
TPM 2.0 Infineon SLB 9670VQ2.0
Operating range
-20 to +60 °C Fanless aluminum · IP40
Form factor
100 × 100 × 45 mm Approx. 457 g · DIN / wall mount

Key capabilities

What separates ALPON X5 AI from a generic edge AI box: cellular connectivity, hardware root of trust, and fleet management are included by default — not sold as add-ons.

01 25 TOPS dedicated AI acceleration

The DEEPX DX-M1 M.2 2280 module delivers 25 TOPS at INT8 on a dedicated NPU with on-chip memory. AI workloads run on the accelerator and do not contend with the host CPU or system RAM. Models in ONNX format are compiled to DXNN and deployed via the dxrt-runtime.

02 Compute Module 5 host

Built on the Raspberry Pi Compute Module 5: a Broadcom BCM2712 quad-core Cortex-A76 at 2.4 GHz, with 8 or 16 GB of LPDDR4-4267 ECC SDRAM. Native compatibility with the Raspberry Pi ecosystem, Debian-based ALPON X5 AI OS, and standard ARM64 container images.

03 Always-online connectivity

2× Gigabit Ethernet, dual-band Wi-Fi 5 (2.4 / 5 GHz), Bluetooth 5.0 / BLE, and LTE Cat 4 with integrated eSIM (eUICC SGP.22). Multi-network support with automatic failover keeps devices reachable even when a single uplink degrades. Integrated GNSS/GPS and cellular geolocation.

04 Hardware root of trust

Onboard Infineon SLB 9670VQ2.0 TPM 2.0 module for secure key storage and measured boot. A hardware watchdog monitors device operation and resets the system on failure, supporting unattended industrial deployments.

05 Industrial form factor

Fanless full-aluminum enclosure with passive top-side cooling, 100 × 100 × 45 mm at approximately 457 g, IP40 rated. Operating range -20 °C to +60 °C, storage -40 °C to +85 °C, up to 95 % relative humidity non-condensing. Optional DIN rail and wall-mount kits.

06 Managed by ALPON Cloud

Zero-touch onboarding, fleet inventory, real-time monitoring, OTA firmware and OS updates, container deployment, and remote terminal access — all in one console. The same platform manages ALPON X4 and ALPON X5 AI fleets side by side.

Where ALPON X5 AI fits

Workloads where on-device inference, cellular failover, and remote management combine into a deployable system.

Computer vision & analytics

  • Real-time object detection (YOLO family)
  • People counting and traffic analytics
  • Defect detection on production lines
  • License plate recognition (ALPR)
  • Smart camera and CCTV analytics

Robotics & autonomy

  • Autonomous mobile robots (AMRs)
  • Vision-based navigation
  • Edge inference for ROS / ROS 2
  • Multi-stream sensor fusion
  • Predictive maintenance

Industrial AI infrastructure

  • On-device inference plus telemetry forwarding
  • Modbus and OPC-UA aggregation with AI scoring
  • Anomaly detection on machine sensor streams
  • PLC supervision with vision feedback
  • Energy and asset condition monitoring

Smart spaces & retail

  • Footfall and occupancy analytics
  • Queue management and dwell-time tracking
  • Retail shelf monitoring
  • Access control with face or plate verification
  • Digital signage with audience-aware content

Platform & ecosystem

ALPON X5 AI is part of a managed edge platform. The same console, container registry, and SIM management work across ALPON X4 and ALPON X5 AI deployments.

Managed by ALPON Cloud

Zero-touch onboarding, fleet inventory, OTA updates, remote terminal, container deployment, and real-time monitoring — for X4 and X5 AI fleets in one place.

Explore ALPON Cloud

Choosing between ALPON X5 AI and ALPON X4

If you do not need on-device AI inference, the ALPON X4 is the cost-effective option for IoT gateway and telemetry workloads. Side-by-side comparison covers compute, memory, ports, and intended workloads.

Compare devices

Start here

Common entry points into the ALPON X5 AI documentation.

Quickstart Power on the device, register it to ALPON Cloud, and reach the dashboard in 5 steps. Physical Layout & I/O Front and rear panel diagrams, port pinouts, button and LED behavior. Power & Mounting Power input options, current draw, DIN rail and wall mount installation. Specifications Full reference: compute, memory, I/O, cellular, power, mechanical, compliance. ALPON X5 AI OS First boot, default credentials, SSH access, OTA updates, system architecture. AI Development Sixfab Model Zoo, ONNX → DXNN compilation, dxrt-runtime, custom models. Safety & Compliance CE, FCC, RoHS, REACH declarations, RF exposure, and installation safety. Troubleshooting Diagnostics for boot, NPU, connectivity, OS, and common deployment issues. FAQ Common questions about deployment, accessories, and X4 compatibility. Connectivity & Antennas LTE bands, antenna placement, GNSS, Wi-Fi best practices for industrial sites. Deploy Applications Push Docker workloads to one device or your entire fleet, with safe rollback. Security Best Practices TPM-backed identity, secure boot, network hardening, and credential rotation.

Updated 5 days ago