eIQ® Inference with Arm® NN

eIQArmNN

eIQ software supports the Arm NN SDK – an inference engine framework that provides a bridge between neural network (NN) frameworks and Arm machine learning processors, including NXP’s i.MX and Layerscape® processors. For Arm® Cortex®-A based processors, Arm NN converts models trained with existing neural network frameworks into inference engines that leverage Arm Neon™ instructions via the Arm Compute Library. NXP provides back-end support for Arm NN to allow inferencing to be run on the i.MX 8 GPUs and the i.MX 8M Plus NPU. Arm NN works with Caffe, TensorFlow, TensorFlow Lite, and ONNX models.

Software Details

Select a section:

Diagram

eIQ Software for Arm NN SDK

eIQ Software for Arm NN SDK

Features

  • Inference engine for Arm Cortex-A CPU cores, Verisilicon GPUs and NPU (i.MX 8M Plus)
  • Support for Caffe, TensorFlow, TensorFlow Lite, and ONNX models
  • Enablement for models such as MobileNet SSD, DeepSpeech v1, and segmentation networks
  • Arm NN fully integrated into Yocto BSP supporting i.MX 8
  • More information available at Arm NN

Supported Devices

  • i.MX8MNANO: i.MX 8M Nano Family - Arm® Cortex®-A53, Cortex-M7
  • i.MX8M: i.MX 8M Family - Arm® Cortex®-A53, Cortex-M4, Audio, Voice, Video
  • i.MX8X: i.MX 8X Family – Arm® Cortex®-A35, 3D Graphics, 4K Video, DSP, Error Correcting Code on DDR
  • i.MX8: i.MX 8 Family – Arm® Cortex®-A53, Cortex-A72, Virtualization, Vision, 3D Graphics, 4K Video
  • i.MX8MMINI: i.MX 8M Mini - Arm® Cortex®-A53, Cortex-M4, Audio, Voice, Video

Downloads

Filters

1 download

  • BSP, Drivers and Middleware

    eIQ Machine Learning Software for i.MX Linux 4.14.y

    EXTERNALRev 0Jun 24, 20191 KBEIQ-MLSOFTWARE
    Download

Note: For better experience, software downloads are recommended on desktop.

Design Resources

Select a section:

Documentation

Quick reference to our documentation types.

Filters

1 documents

Hardware

Filters

3 hardware offerings

Related Software

Filters

1 software file

Note: For better experience, software downloads are recommended on desktop.

Training

Filters

1 trainings