Pico-ITX POE Edge AI Embedded Board with NXP i.MX8M Mini ARM Processor, Dual-core DSP for Voice Control, POE, MIPI, LVDS and Edge TPU Support.
The EMB-2237-AI is a Pico-ITX (100x 72mm) POE edge AI embedded board based on NXP i.MX8M Mini ARM application processor. The board features a Power over Ethernet (PoE) Ethernet port, on-board dual-core DSP that runs algorithms for voice control, noise suppression, and echo cancellation technology, a full set of I/Os including RS-232/485, and m.2 PCIe slot for Edge TPU AI based solutions.
Built for Edge AI Applications
Targeted to satisfy diverse applications demanding a robust and compact computing POE edge AI embedded board for industrial machine learning, embedded vision and inferencing, the board is based on NXP application processors with 10+ year longevity support, built-in reliable communication interfaces like RS-485, RS-232 and M.2 slot for support of TPU, VPU accelerator modules like Google Coral Edge TPU and others.
PoE Power Module
Provides on-board Power over Ethernet (PoE) IEEE 802.3af standards power input via RJ45 Ethernet port for the board and system to make installing or large roll-out much simpler and cheaper
Single Point Touch Screen Connector
Integrates 4-lane MIPI DSI, dual Channel LVDS, RGB, power, backlight control and touch panel interface all in one FPC connector to give developers fast, convenient and simplified solutions for industrial and embedded POE touch panel PC applications.
On-board Voice Control Support
Built-in smart codec with dual-core DSP that runs algorithms for voice control, omni-directional spatial, noise suppression, and acoustic echo cancellation (AEC). Linux supports for Amazon AVS (Alexa Voice Service) Device SDK and Sensory TrulyHandsfree Wake Word Engine. Additional 8 channel digital audio/DMIC inputs (SAI5) for MIC array is also available via 40-pin expansion header.
Touch Panel Development Kit
5”, 7” and 10″ touch panel development kits built on the EMB-2237-AI POE edge AI embedded board to help embedded software and application development and speed up time-to-market.
*Specifications are representative. Specific product technology may be different than indicated.
|Processor||NXP i.MX 8M Mini, 4x ARM Cortex-A53 @ 1.8GHz, 1x Arm Cortex-M4 @ 400MHz|
|System Memory||Onboard 2GB LPDDR4, up to 4GB|
|Ethernet||1 x 10/100 Mbps Ethernet with POE support|
|Storage||Onboard iNAND flash (8GB default), 1x micro SD slot, 265K EEPROM with write protect control|
|Wireless||1x WiFi/Bluetooth (AP6212A) module, 802.11 b/g/n + Bluetooth v4.0|
|AI/ML||OpenCL CPU: 32 GOPS; Google Edge TPU coprocessor PCIe option, up to 4 TOPS|
|LCD Interface||1x MIPI 4-lane DSI, 1x Dual Channel LVDS and 24bit RGB for 5”, 7”, 8”, 10.1” and other size LCD panels|
|Display Resolution||Up to 1920x 1200|
|Graphic Controller||GC NanoUltra (1 shader), OpenGL ES 2.0; GC520L 2D|
|Audio||1 x mono Class D speaker out, 2W(4Ω), 2 (R/L) x line-out and amplified (2x 1.5W) out headers|
|Voice Control||Dual digital MEMS microphone header via CS47L24 with dual DSP, support multi-mic noise suppression, acoustic echo cancellation (AEC), omni-directional spatial|
|Expansion||1x POE (Power over Ethernet) IEEE 802.3af Module, 1x M.2-2230 (E key) slot|
|I/O||1x RS-232/RS-485 pin header, 2x USB2.0 Type A, 2x USB2.0 header, 1x USB Type C OTG, 4+ GPIO, 2x I2C (for TP and MIPI CSI), 1x MIPI 4-lane CSI|
|OS Support||Embedded Linux (kernel version 4.14, Qt and Wayland), Amazon AVS (Alexa Voice Service) Device SDK, Sensory TrulyHandsfree Wake Word Engine, Android 9|
|Watchdog timer||Programmable timer system reset|
|Power Input||12V DC barrel jack or POE via RJ45|
|Form Factor||Embedded Pico-ITX standard|
|Operating Temperature||0 ~ 60°C (32 ~ 140°F)|
|Storage||-40 ~ 85° C (-40 ~ 185° F)|
|Operating Humidity||5% ~ 95% @ 40°C (Non-Condensing)|
|Dimension||100 x 72 mm (3.94” x 2.83”)|