What are the Arm NN SDK and the Arm NN APIs?

Arm NN SDK is a set of open-source Linux software tools that enables machine learning workloads on power-efficient devices. In the application that we use in this guide, we use Arm NN to enhance performance on the Arm architecture. The inference engine provides a bridge between existing neural network frameworks and Arm Cortex-A CPUs, Arm Mali GPUs, and NPUs.

Android Neural Networks API and TensorFlow Lite delegate

The Android Neural Networks API (NNAPI) is an Android C API that is designed to run computationally intensive operations for Machine Learning on mobile devices.

NNAPI supports various hardware accelerations. NNAPI uses TensorFlow as a core technology. If you build your mobile app with TensorFlow, your app gets the benefits of hardware acceleration through the NNAPI. NNAPI abstracts the hardware layer for ML inference. Arm NN works with Android NNAPI to target Arm processors, enabling exponential performance boosts.

A TensorFlow Lite delegate is a way to delegate part or all of graph execution to another executor. Running computationally intensive NN models on mobile devices is resource demanding on mobile CPUs. This means that devices with hardware accelerators provide better performance and higher energy efficiency through NNAPI.

TensorFlow Lite uses an NNAPI delegate to access NNAPI. You can access their open source code.

Previous Next