We have various microcontroller and microprocessor boards available in the market. Not all boards support TensorFlow Lite. Rather then saying that, not all boards have TensorFlow prepared and tested. We are talking about the microcontrollers here, so let us list the microcontroller boards, also, let us look at embedded boards which support TensorFlow (Lite).
| MCU / Board | Processor | Framework | Notes |
|---|---|---|---|
| Arduino Nano 33 BLE Sense | nRF52840 (ARM Cortex-M4F) | TFLite Micro | Popular for audio/sensor models |
| Arduino Nicla Vision / Nicla Sense ME | STM32H747 / nRF52832 | TFLite Micro | Vision & sensor fusion |
| Arduino Uno R4 WiFi | Renesas RA4M1 (Arm Cortex-M4F) | TFLite Micro | Supported via TinyML + TFLM |
| ESP32 / ESP32-S3 / ESP32-C6 | Xtensa / RISC-V MCU | TFLite Micro | ESP-NN acceleration for int8 ops |
| STM32F4 / F7 / H7 series | ARM Cortex-M4/M7 | TFLite Micro | ST Cube.AI or STM32 AI tools integration |
| SparkFun Edge | Ambiq Apollo3 (ARM Cortex-M4F) | TFLite Micro | Ultra-low-power TinyML reference board |
| Adafruit EdgeBadge / Feather nRF52840 | nRF52840 | TFLite Micro | Works with Arduino + TensorFlow Lite libraries |
| Seeed Wio Terminal | ATSAMD51 + RTL8720DN | TFLite Micro | Compatible with TensorFlow Lite Arduino library |
| Kendryte K210 boards (e.g., Maix Bit, Maix Dock) | Dual RISC-V AI cores | Vendor SDK (not TFLM) | Runs pre-compiled TFLite models using KPU |
]]>
0 Comments