The Grinn GenioBoard is a compact single-board computer (SBC) designed to shorten the development cycle of embedded and AI-enabled edge devices.
It is built around MediaTek Genio series processors and integrates either the GenioSOM-510 or GenioSOM-700 compute module.
Both SOM modules are pin-compatible, enabling developers to switch between performance tiers without redesigning the carrier board, significantly reducing hardware revision effort and cost.
Performance: Balanced Compute and Power Efficiency
(1) Processor Architecture
GenioSOM-700: Powered by MediaTek Genio 700
- 8-core CPU: 4× Arm Cortex-A78 + 4× Cortex-A55
- Optimized for edge inference and higher compute demands.
GenioSOM-510: Powered by MediaTek Genio 510
- 6-core CPU: 2× Arm Cortex-A78 + 4× Cortex-A55
- Offers a balanced ratio of performance and efficiency for mid-range designs.
(2) AI and Graphics Acceleration
Both SOMs integrate:
- NPU (Neural Processing Unit) for real-time edge inference
- GPU for hardware-accelerated graphics and vision workloads
Suitable for computer vision, video analytics, and local inference where latency matters.
GenioBoard-700 Block Diagram
Hardware Features: Compact, Expandable, Developer-Friendly
(1) Memory and Storage
- Up to 8GB LPDDR4/4X
- On-board 16GB eMMC
- M.2 M-Key slot (PCIe Gen2 x1): supports NVMe SSD or AI accelerator
- Future expansion options include additional eMMC or UFS via SOM configuration
(2) Connectivity and I/O Interfaces
Display: HDMI + DisplayPort dual output
Cameras: 2× MIPI-CSI (4-lane each)
Networking: Gigabit Ethernet (RGMII PHY)
Wireless expansion: M.2 E-Key for Wi-Fi / Bluetooth modules
USB:
- 1× USB-C (power input, 5–20V USB-C PD)
- 1× USB-C (debug)
- 2× USB 3.0 Type-A
- 2× USB 2.0 Type-A
GPIO:
- 40-pin Raspberry Pi–compatible HAT header
- Supports UART / I²C / SPI / PWM
- 4-pin fan connector for cooling
(3) Form Factor and Security
Compact footprint: 87 × 56 mm
Integrated Infineon OPTIGA Trust M security module
- Secure key storage
- Device attestation
- Secure boot support (CRA-standard compliant)
Software Ecosystem
Supports Yocto Project Linux and Debian Linux
Android support under development
Open-source resources available:
- Developer documentation and wiki
- GitHub repository (meta-grinn-genio) — MIT license
- BSP build scripts and Docker environment for reproducible builds
Core Specification Table
| Category | GenioSOM-510 | GenioSOM-700 |
|---|---|---|
| Processor (SoC) | MediaTek Genio 510 6 cores (2× Cortex-A78 + 4× Cortex-A55) |
MediaTek Genio 700 8 cores (4× Cortex-A78 + 4× Cortex-A55) |
| AI / Graphics | Integrated NPU + GPU, supports edge AI inference and HW-accelerated graphics | |
| Memory / Storage | Up to 8GB LPDDR4/4X 16GB eMMC |
Up to 8GB LPDDR4/4X 16GB eMMC |
| Storage Expansion | M.2 M-Key (PCIe Gen2 ×1) for NVMe SSD / AI accelerator Optional eMMC / UFS expansion via SOM |
|
| Display / Camera Interfaces | HDMI + DisplayPort 2× 4-lane MIPI-CSI |
|
| Connectivity | Gigabit Ethernet (RGMII PHY) M.2 E-Key for Wi-Fi / Bluetooth 40-pin Raspberry Pi HAT header |
|
| USB Interfaces | 1× USB-C (power input 5–20V PD) 1× USB-C (debug) 2× USB 3.0-A + 2× USB 2.0-A |
|
| Security | Infineon OPTIGA Trust M, secure key storage & secure boot | |
| Dimensions | 87 × 56 mm | |
GenioBoard is suitable for:
- Industrial IoT monitoring
- Smart retail terminals
- Edge AI security cameras
- Drones or robotics with on-device inference
At the time of writing, pricing has not yet been publicly released, but the product and technical documentation are available on the official Grinn website.
