Efinix released the TinyML platform for AI acceleration on its FPGAs. The platform works with the high performance Sapphire RISC-V processor. Comprising a model profiler and GUI for accelerator selection, the platform delivers an accelerated AI capability to the quad-core-capable Sapphire core. It leverages the RISC-V custom instruction capability to deliver a scalable library of accelerated instructions customized for the requirements of the chosen TensorFlow Lite model. Instructions can be selected through a GUI to deliver the resource / acceleration tradeoff needed.
