Nov 5, 2025 — by Giorgio Mendoza 225 views

Grinn has unveiled the GenioBoard, a compact single-board computer aimed at accelerating development of embedded and AI-enabled systems. It integrates the company’s GenioSOM-510 and GenioSOM-700 modules built on MediaTek’s Genio processor family, combining multiple Arm Cortex-A cores with an integrated GPU and NPU for edge inference applications.

Powered by the MediaTek Genio 700, which integrates a quad-core Arm Cortex-A78 and dual-core Cortex-A55 CPU cluster, the GenioBoard delivers high performance with low power consumption. The system also includes an integrated GPU and a dedicated AI processing unit for on-device inference.

— ADVERTISEMENT —

The GenioSOM-510 variant offers a lower-power configuration based on the MediaTek Genio 510 (2× Cortex-A78 + 4× Cortex-A55), while the GenioSOM-700 features an eight-core CPU (2× Cortex-A78 + 6× Cortex-A55). Both are pin-compatible, allowing developers to scale between performance tiers without altering the carrier layout.

GenioBoard-700 Block Diagram(click images to enlarge)

Memory options extend up to 8 GB LPDDR4, paired with onboard eMMC storage and an M.2 M slot for SSD or AI accelerator expansion. The GenioBoard is available in configurations with 4 GB RAM and 16 GB eMMC.

GenioBoard-700 Bottom View(click images to enlarge)

The board measures 87 × 56 mm and exposes key interfaces from the SOMs, including dual CSI-2 camera inputs, HDMI and DisplayPort outputs, USB 3.0/2.0, and Gigabit Ethernet with RGMII PHY.

GenioBoard-700 Pinout(click images to enlarge)

Expansion options include M.2 Key M for accelerators and M.2 Key E for Wi-Fi or Bluetooth connectivity, while a Raspberry Pi HAT-compatible header supports additional peripherals.

Grinn GenioBoard-700(click images to enlarge)

According to the company, the GenioBoard supports Yocto-based and Debian Linux distributions, with Android support currently in development.

Documentation is available through a developer wiki, and Grinn maintains an open-source meta-grinn-genio repository on GitHub under the MIT license, including build and flashing guides, Docker image scripts, and BSP layers for the GenioSOM modules.

Specifications listed for the Grinn GenioBoard include:

• Processor/SoC:

• MediaTek Genio 700 with 8 cores (4 × Arm Cortex-A78 + 4 × Cortex-A55)

• Optional Genio 510 with 6 cores (2 × Cortex-A78 + 4 × Cortex-A55)

• Integrated NPU

• Graphics/AI Acceleration:

• Integrated GPU w/ hardware-accelerated graphics & vision processing

• Dedicated AI engine for edge inference workloads

• 1× M.2 E-Key slot for external AI accelerator modules

• Memory/Storage:

• Up to 8 GB LPDDR4 RAM

• 16 GB onboard eMMC flash (expandable)

• 1× M.2 M-Key PCIe x1 Gen2 slot for SSD or AI accelerator

• Support for additional eMMC & UFS modules via SOM

• Display:

• 1× HDMI output

• 1× DisplayPort (DP) output

• Camera:

• 2× 4-lane MIPI-CSI interfaces

• Connectivity:

• 1× Gigabit Ethernet PHY (RGMII)

• 2× USB-A 3.0 ports

• 2× USB-A 2.0 ports

• 1× USB-C to UART interface

• 1× Raspberry Pi HAT-compatible 40-pin GPIO header

• M.2 E-Key slot for Wi-Fi and Bluetooth modules

• Security:

• Infineon OPTIGA Trust M SLS32AIA010M hardware security module

• Hardware-based key storage & cryptographic operations

• CRA-ready secure boot & authentication support

• I/O Interfaces:

• 40-pin GPIO header (UART, I²C, SPI, GPIO, PWM)

• 4-pin fan connector

• USB:

• 1× USB Type-C (Power Delivery input, 5–20 V)

• 1× USB Type-C (debug/programming interface)

• 2× USB 2.0 Type-A

• 2× USB 3.0 Type-A

• Power:

• Single USB Type-C Power Delivery (PD) input (5 – 20 V)

• Mechanical:

• 87 mm × 56 mm

Further Information

Grinn has not yet announced pricing, but the GenioBoard product page is live on the company’s website.