
1 / 4 · Click to enlarge
AI Summary
A compact AIoT module featuring the dual-core RISC-V K210 processor with 8MB SRAM and dedicated neural and audio processing. It enables local machine vision and hearing in a tiny 25 x 25 mm footprint, without WiFi.
Key Highlights
- Dual-core RISC-V 64-bit CPU adjustable to 400MHz
- 8MB on-chip SRAM split for general and AI use
- Neural network KPU and high-performance audio processor
- FPIOA maps 255 functions to 48 GPIOs
- DVP camera and MCU LCD interfaces
- Supports FreeRTOS, standard SDK and MicroPython on M1
Ideal For
- Embedded engineers building edge AI prototypes
- Makers exploring RISC-V and machine vision
- Product developers needing compact AI modules
- Researchers working on machine hearing
- IoT designers creating offline smart devices
Buying Tip
Remember that this version has no WiFi, so plan for wired or offline operation, and budget for a separate DVP camera and LCD if your project needs them.
Elecvora expert review
Our verdict
The Sipeed MAIX-I packs a remarkable amount of AI capability into a tiny module, making it a strong choice for edge vision and audio projects. However, the lack of WiFi and the need for external peripherals mean it is best for developers who are comfortable with embedded workflows.
AI-assisted analysis based on the manufacturer's specifications. Not a hands-on lab test.
Pros
- Powerful RISC-V dual-core CPU with dedicated AI and audio processors
- Large 8MB on-chip SRAM reduces need for external memory
- Flexible FPIOA simplifies custom carrier board design
- Compact 25 x 25 mm size and wide operating temperature range
Cons
- No WiFi connectivity in this version
- Requires separate camera and display modules for full functionality
- Limited documentation compared to more established platforms
Best for
Embedded developers and makers building offline AI vision or audio devices
Not ideal for
Users needing wireless connectivity or a complete plug-and-play solution
Specifications
| Item Type | Sipeed Maix |
|---|---|
| Processor Type | RISC-V Dual Core 64bit |
| CPU | RISC-V Dual Core 64bit, 400Mh adjustable |
| FPU | IEEE754-2008 compliant high-performance pipelined FPU |
| Debugging Support | High-speed UART and JTAG interface for debugging |
| SRAM | The SRAM is split into two parts, 6MiB of on-chip general-purpose SRAM memory and 2MiB of on-chip AI SRAM memory, for a total of 8MiB |
| FPIOA | FPIOA allows users to map 255 internal functions to 48 free IOs on the chip, Field Programmable IO Array (FPIOA/IOMUX) |
| Digital Video Port (DVP) | Max. frame size 640x480 |
| FreeRtos & Standard SDK | Supports FreeRtos and Standard development kit. |
| MicroPython Support | Support MicroPython on M1 |
| Machine vision | Machine vision based on convolutional neural network |
| Machine hearing | High performance microphone array processor |
| Supply Voltage | 5.0 ±0.2 |
| Supply Current | & 300 |
| Length (mm) | 25 |
| Width (mm) | 25 |
| Height (mm) | 1 |
| Weight (g) | 8 |
| Operating Temperature (°C) | -30℃ ~ 85℃ |
| Shipping Weight | 0.02 kg |
| Shipping Dimensions | 3 × 3 × 2 cm |
Frequently Asked Questions
Does this module include WiFi?
What is the power supply requirement?
What software development options are supported?
Can I connect a camera and display?
What is included in the package?
What is the operating temperature range?
Overview
The Sipeed MAIX-I is an AIoT module designed to bring machine learning to the edge. At its heart is the Kendryte K210, a dual-core RISC-V 64-bit processor with a floating-point unit and a dedicated neural network processor. This version is supplied without WiFi, so it is best suited to standalone or wired applications where wireless connectivity is not required. The module measures only 25 x 25 x 1 mm and weighs 8 grams, making it easy to integrate into space-constrained designs.
Key Features
- RISC-V dual-core CPU: 64-bit cores with adjustable frequency up to 400MHz and an IEEE754-2008 compliant FPU.
- Generous on-chip memory: 6MiB general-purpose SRAM plus 2MiB AI SRAM for a total of 8MiB.
- AI acceleration: A neural network processor supports convolutional neural networks for machine vision tasks.
- Audio processing: A high-performance microphone array processor enables machine hearing applications.
- Flexible IO: The Field Programmable IO Array (FPIOA) lets you map up to 255 internal functions to 48 free IOs.
- Camera and display interfaces: A DVP port supports frame sizes up to 640x480, and an MCU LCD interface is available for display output.
- Development support: High-speed UART and JTAG interfaces are provided for debugging, and the module supports FreeRTOS, a standard SDK, and MicroPython on M1.
Specifications Explained
The 400MHz default clock speed can be adjusted to balance performance and power. The 8MiB SRAM is unusually large for a microcontroller-class device, allowing AI models and buffers to reside on-chip without external memory. The FPIOA is particularly useful for custom carrier boards, as it allows peripheral functions to be routed to almost any pin. The DVP interface accepts common camera modules, while the LCD interface simplifies adding a small display for preview or UI. The module operates from a 5.0V ±0.2V supply and draws up to 300mA, with an operating temperature range of -30°C to 85°C.
Compatibility and Requirements
The MAIX-I requires a 5V power source capable of supplying at least 300mA. Debugging is possible via UART or JTAG. Software development can be done with FreeRTOS and the standard SDK, or with MicroPython on M1. No WiFi functionality is included in this version. A DVP camera and MCU LCD are optional and must be sourced separately if needed.
Typical Uses
- Machine vision projects such as object recognition and image classification
- Machine hearing applications using microphone arrays
- Edge AI prototyping and proof-of-concept devices
- Robotics and automation controllers
- Smart sensors and IoT endpoints without wireless connectivity
What's in the Box
Package contents are not specified by the manufacturer. Please check the product listing for details.

Quick specs
- Item Type
- Sipeed Maix
- Processor Type
- RISC-V Dual Core 64bit
- CPU
- RISC-V Dual Core 64bit, 400Mh adjustable
- FPU
- IEEE754-2008 compliant high-performance pipelined FPU
- Debugging Support
- High-speed UART and JTAG interface for debugging
- SRAM
- The SRAM is split into two parts, 6MiB of on-chip general-purpose SRAM memory and 2MiB of on-chip AI SRAM memory, for a total of 8MiB



















