ProductTFLite Micro
TensorFlow Lite for Microcontrollers
A lightweight inference engine released by Google for microcontroller scenarios. It provides a streamlined interpreter, fixed-point and quantized operator libraries, and static memory planning capabilities, requires no full C++ standard library, and has long been the de facto standard deployment solution on Cortex-M.
Timeline (last 90 days)
Nov 8
On Cortex-M, INT8 quantization is almost the default option because most MCUs lack an efficient floating-point unit; on the toolchain side, TensorFlow Lite for Microcontrollers, Arm CMSIS-NN, and the Ethos-U Vela compiler together form the chain from model conversion to operator optimization.
Unverified50%