Specifications
Model
Banana Pi BPI-AI2N (BPI-AI2N-Core SoM plus carrier board)
Product Type
System-on-module with a Jetson-Nano-style carrier board for Vision AI applications
SoC
Renesas RZ/V2N
CPU
Quad-core Arm Cortex-A55 @ 1.8 GHz
GPU
Arm Mali-G31 3D graphics engine (GE3D) — OpenGL ES 3.2 and OpenCL 2.0 FP
AI Accelerator
DRP-AI3 — up to 4 dense TOPS or 15 sparse TOPS
H.264
Up to 1920 x 1080 @ 60 fps
H.265
Up to 3840 x 2160 @ 30 fps
System Memory
8 GB LPDDR4X
Storage
32 GB eMMC flash plus 64 MB SPI flash
Networking (SoM)
2x Realtek RTL8211F Gigabit Ethernet controllers
SoM Connector
260-pin SO-DIMM edge connector exposing all I/O
Power (SoM)
On-module PMIC
Carrier — Storage
M.2 Key-M 2280 socket for NVMe SSD
Carrier — Display
MIPI DSI connector
Carrier — Cameras
2x MIPI CSI connectors
Carrier — Networking
2x Gigabit Ethernet RJ45 ports
Carrier — Wireless
Wi-Fi 5 and Bluetooth 4.2 via an RTL8821CU USB module
Carrier Form Factor
Jetson Nano-inspired layout
Overview
The Banana Pi BPI-AI2N pairs a system-on-module built on the Renesas RZ/V2N with a carrier board laid out along the lines of a Jetson Nano. The split matters commercially: the 260-pin SO-DIMM module carries the SoC, memory, storage and PMIC, so a product can be designed once around the module and reworked later without redesigning the baseboard.
Vision AI is handled by Renesas's DRP-AI3 dynamically reconfigurable processor, rated at up to 4 dense TOPS or 15 sparse TOPS. That sits alongside a quad-core Cortex-A55 cluster at 1.8 GHz, an Arm Mali-G31 GPU with OpenGL ES 3.2 and OpenCL 2.0 FP, H.264 decode up to 1080p60 and H.265 up to 4K30. The module carries 8 GB of LPDDR4X, 32 GB of eMMC and 64 MB of SPI flash, and its two Realtek RTL8211F Gigabit controllers are brought out as dual RJ45 on the carrier.
The carrier board adds an M.2 Key-M 2280 socket for NVMe storage, a MIPI DSI display connector, two MIPI CSI camera connectors, dual Gigabit Ethernet, and Wi-Fi 5 with Bluetooth 4.2 through an RTL8821CU USB module. The Jetson Nano-derived mechanical layout means existing enclosures, camera brackets and cabling designed around that footprint are largely reusable, which is the practical argument for choosing this module over another RZ/V2N board.
Key Benefits
DRP-AI3 accelerator at 15 TOPS sparse targets Vision AI inference on a low-power, low-cost RZ/V2N platform. 260-pin SO-DIMM module with on-module PMIC separates the design cycle from the manufacturing cycle. 8 GB LPDDR4X, 32 GB eMMC and 64 MB SPI flash are populated on the module, so the carrier needs no memory design. M.2 2280 NVMe plus dual MIPI CSI and dual Gigabit Ethernet on a Jetson-Nano-style layout makes it a drop-in alternative for existing vision enclosures.
Applications
Vision AI cameras and inspection systems, industrial robotics and AGV perception, smart-city and access-control vision, retail analytics endpoints, replacement of Jetson Nano class boards where a Renesas DRP-AI pipeline is preferred.
Request a Quote — BANANA PI BPI-AI2N — RENESAS RZ/V2N SOM AND CARRIER BOARD FOR VISION AI
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →