AI

The **ROM-5722WD-RFA1E** is a high-performance SMARC (Smart Mobility ARChitecture) 2.1 module manufactured by **Advantech**. It is built on the NXP i.MX 8M Plus platform, designed specifically for AI-at-the-edge, machine learning, and industrial multimedia applications.
---
## 1. Core Technical Specifications
The following table highlights the primary electronic components and processing capabilities of the module.
| Feature | Specification |
| :--- | :--- |
| **Processor** | NXP i.MX 8M Plus Quad (4x Arm Cortex-A53 @ 1.8 GHz) |
| **Co-Processor** | Arm Cortex-M7 (800 MHz) for real-time processing |
| **NPU (AI Accelerator)** | Integrated Neural Processing Unit (up to 2.3 TOPS) |
| **Memory (RAM)** | 4GB LPDDR4 (Industrial Grade) |
| **Storage (eMMC)** | 32GB eMMC (On-board flash) |
| **Form Factor** | SMARC 2.1 Standard (82 x 50 mm) |
| **Operating Temp** | -40 ~ 85°C (Wide Temperature/Industrial Grade) |
---
## 2. Key Electronic Functional Blocks
### A. AI and Vision Processing
The standout electronic component is the **Integrated NPU**. This allows the module to handle complex machine learning tasks (like object detection and speech recognition) locally without relying on a cloud connection. It also features a Dual Image Signal Processor (ISP) for high-quality camera input.
### B. Graphics and Multimedia
The module integrates a Vivante GC7000UL GPU, supporting:
* **Video Decode:** 1080p60 HEVC/H.265, H.264, VP9, VP8.
* **Video Encode:** 1080p60 H.265, H.264.
* **Display Interfaces:** Dual-channel LVDS (or MIPI-DSI) and HDMI 2.0.
### C. Connectivity and I/O
As a SMARC module, it routes signals through a 314-pin edge connector:
* **Networking:** Dual Gigabit Ethernet (one with TSN - Time Sensitive Networking support).
* **USB:** 1x USB 3.0, 2x USB 2.0.
* **PCIe:** 1x PCIe Gen 3 slot for expansion (e.g., Wi-Fi or NVMe).
* **Industrial I/O:** CAN-FD, UART, I2C, SPI, and GPIO.
---
## 3. Part Number Decoder: ROM-5722WD-RFA1E
Advantech part numbers are structured to define specific hardware configurations:
* **ROM-5722:** The base model series (i.MX 8M Plus SMARC).
* **W:** Wide Temperature support (-40 to 85°C).
* **D:** Dual/Quad Core configuration (Quad in this specific variant).
* **RFA1E:** Revision code and specific packaging/firmware version.
---
## 4. Power Requirements
The module is designed for low power consumption, typically drawing between **4W to 6W** depending on the NPU load. It requires a standard **5V DC** input via the SMARC carrier board.
- ⤷
What are the compatible carrier boards for the ROM-5722 series?
- ⤷ How do I implement the NPU for TensorFlow Lite models on this module?
- ⤷ What is the difference between SMARC 2.1 and previous SMARC standards for this hardware?