All News | Boards | Chips | Devices | Software | LinuxDevices.com Archive | About | Contact | Subscribe
Follow LinuxGizmos:
Twitter Facebook Pinterest RSS feed
*   get email updates   *

ArmSoM RK3588 AI Module7 Launches as Jetson Nano-Compatible Edge Platform

Aug 21, 2025 — by Giorgio Mendoza 214 views

First unveiled a year ago, ArmSoM has finally launched the RK3588 AI Module7 (AIM7) on Crowd Supply. The open hardware compute module is designed as a Jetson Nano–compatible alternative, targeting edge computing, vision systems, and embedded AI projects.

The module is powered by the Rockchip RK3588, an octa-core processor combining four Cortex-A76 performance cores at up to 2.4 GHz with four Cortex-A55 efficiency cores at 1.8 GHz. A Mali-G610 MP4 GPU provides graphics acceleration with support for Cortex-A55 and Vulkan 1.4, while a built-in NPU delivers up to 6 TOPS of AI performance for machine learning inference at the edge.

— ADVERTISEMENT —


RK3588 block diagram
(click image to enlarge)

Memory is configured with 8 GB of LPDDR4x, alongside 64 GB or 128 GB of eMMC 5.1 storage and a microSD card slot for expansion. The 260-pin edge connector is mechanically compatible with Jetson Nano carrier boards, while the hardware exposes PCIe 3.0, PCIe 2.0, HDMI 2.1, DP 1.4, and multiple MIPI CSI interfaces for multi-camera setups.

RK3588 AI Module7 Benchmark
(click image to enlarge)

ArmSoM also provides the AIM-IO carrier board, which serves as an expansion platform for the module. It integrates Gigabit Ethernet, four USB 3.0 ports, M.2 expansion for Wi-Fi or 5G modules, dual 4K display outputs, and 40-pin GPIO headers. Multiple MIPI camera connectors and optional PoE support make it a practical platform for AI vision, robotics, or industrial control applications.

 RK3588 AI Module7 & Carrier board
(click image to enlarge)

Software support includes Debian, Ubuntu, and Armbian, with AI optimization handled through the RKNN-Toolkit2. This toolchain enables one-click conversion of TensorFlow and PyTorch models to RKNN format, INT8 quantization for reducing model size, and real-time profiling for performance tuning.

Specifications listed for the AIM-IO carrier board:

  • Memory/Storage:
    • Up to 32GB LPDDR4x RAM, 2112 MHz
    • Up to 32GB eMMC 5.1
    • 1x MicroSD card slot
  •  Audio/Display:
    • 1x DisplayPort
    • 1x HDMI-out
    • 1x 4 lanes MIPI DSI up to 4K@60fps
  • Camera:
    • 2x 2 lanes MIPI CSI
    • 1x 4 lanes MIPI CSI
  •  Connectivity:
    • 1x Gigabit Ethernet port
  •  Expansion:
    •  1x M.2 slot (E-key, PCIe/USB/SDIO/UART)
  • I/O Interfaces:
    • 40-pin GPIO expansion header
  • USB:
    • 4x USB 3.0 Type-A
  • Power:
    • 5V (via Barrel jack)
  • Mechanical:
    • 100 x 80 x 29 mm

Further Information

The RK3588 AI Module7 is available on Crowd Supply starting at $239, with shipping expected to begin in January 2026.

(advertise here)


PLEASE COMMENT BELOW

Please comment here...