Skip to content

OPEN. POWERFUL. FOR DEVELOPERS.

Build the Next Generation
of Embedded Devices

μForge.io is an open-source development platform for innovative semiconductor solutions, beginning with the SF32 family of ultra-low-power microcontrollers from SiFli. From silicon, modules, and development boards to SDKs, documentation, and production-ready reference designs, μForge.io brings together everything developers need to design, build, and ship connected, intelligent, and energy-efficient embedded products, faster.

Get Started Explore SF32 GitHub

Open Source     Active Community     Production Ready     Regular Updates

SF32 MCU — Bluetooth 6.3, 2D/2.5D GPU, multiple display interfaces, AI accelerator, ultra-low power, audio codec

  • Hardware

    Microcontrollers, modules, development boards, and production-ready reference designs.

    See Hardware

  • SDK & Tools

    SDKs, toolchains, middleware, drivers, and developer tools.

    View Develop

  • Learn

    Architecture and subsystem guidance for graphics, Bluetooth, AI, audio, and low power.

    Explore Learn

  • Examples

    Production-quality example applications for common embedded use cases.

    Browse Examples

Why Choose SF32?

  • High Performance

    Single-core, dual-core or triple-core Cortex-M33 / STAR-MC1, featuring integrated instruction (I$) and data (D$) caches, FPU and MPU, and up to 64 MB of in-package PSRAM for graphics-intensive and compute-intensive applications.

  • Bluetooth 6.3

    Dual-mode Bluetooth with Bluetooth Classic and LE, LE Audio, and long-range PHY support. A dedicated Bluetooth subsystem with its own processor, memory, and firmware enables advanced customization and protocol development.

  • Graphics Acceleration

    2D / 2.5D GPU, LCD controller, and LVGL graphics framework for rich and smooth graphical user interfaces with resolutions up to 1280 × 720. Supported interfaces include MIPI, RGB, QSPI, 8080, JDI, and EPD.

  • Audio Subsystem

    Integrated audio subsystem with analog audio, I²S, PDM, and hardware-assisted digital audio processing.

  • AI Acceleration

    Integrated NPU for TinyML, voice recognition, and intelligent edge AI applications.

  • Ultra-Low Power

    Advanced power management architecture optimized for battery-powered and always-on devices.

Supported SF32 MCUs 1

  • SF32LB52

    SF32LB52x Essential Wearable MCU

    • Cortex-M33 STAR-MC1
    • Dual-Mode Bluetooth 6.3
    • 2D / 2.5D GPU, ePicasso 2.0
    • eZip 2.0 Hardware Decompression
    • Audio Codec, 1 x AMIC, 2 x DMIC
    • 576 KB SRAM, up to 16 MB PSRAM
    • QFN68 package, with up to 45 GPIOs and optional integrated linear charger

    View Series

  • SF32LB55

    SF32LB55x Mainstream BLE Wearable MCU

    • Cortex-M33 STAR-MC1
    • Single-Mode Bluetooth 5.2
    • 2D / 2.5D GPU, ePicasso 1.0
    • eZip 1.0 Hardware Decompression
    • AI Accelerator
    • 1.4 MB SRAM, up to 18 MB PSRAM
    • Available in QFN68, BGA145, and BGA169 packages, with up to 49 / 95 / 119 GPIOs

    View Series

  • SF32LB56

    SF32LB56x Graphics-Optimized MCU

    • Cortex-M33 STAR-MC1
    • RGB Display up to 1024 x 600
    • 2D / 2.5D GPU, ePicasso 2.0
    • eZip 2.0 Hardware Decompression
    • Audio Codec, 1 x AMIC, 4 x DMIC
    • 960 KB SRAM, up to 16 MB PSRAM
    • Available in QFN68 and BGA175 packages, with up to 44 / 120 GPIOs

    View Series

  • SF32LB58

    SF32LB58x Flagship AIoT Platform MCU

    • Dual-core Cortex-M33 STAR-MC1
    • MIPI Display up to 1280 x 720
    • 2D / 2.5D GPU + Vector Graphics, ePicasso 2.0
    • eZip 2.0 and JPEG/MJPEG Dec
    • Audio Codec, 2 x AMIC, 4 x DMIC
    • 3.7 MB SRAM, up to 64 MB PSRAM
    • Available in BGA256 package, up to 154 GPIOs

    View Series

  1. See the SF32 Family Overview for a detailed comparison of all devices. ↩