eIQ® Inference with Glow NN
The eIQ machine learning (ML) software development environment for i.MX RT crossover MCUs supports the Glow machine learning compiler, which enables ahead-of-time compilation. The compiler converts the neural networks into object files, then the user converts this into a binary image for increased performance and smaller memory footprint as compared to a traditional runtime inference engine.
Glow is used as a software back-end for the PyTorch machine learning framework, including support for the ONNX model format.
Glow, or graph lowering, compiler derives its name because it lowers a neural network into a two-phase strongly typed intermediate representation. In the first phase, the optimizer performs domain-specific optimizations. The second phase allows the compiler to perform optimizations that take advantage of specialized back-end hardware features. It’s in this second phase that NXP has added specialized support for Arm® Cortex®-M cores and Cadence® Tensilica® HiFi 4 DSP support, accelerating performance by utilizing Arm CMSIS-NN and HiFi NN libraries, respectively.
Software Details
Diagram
Features
- Support for i.MX RT600, i.MX RT1050, i.MX RT1060, and other crossover MCUs
- Enables inference at the edge with significantly lower latency and smaller binary size
- Utilizes CMSIS-NN and Cadence® HiFi NN Library to accelerate neural network operators
- Delivered within MCUXpresso SDK
- Glow is an open source, community project released from Facebook in 2018 with Apache License 2.0
Supported Devices
- i.MX-RT600: i.MX RT600 Crossover MCU with Arm® Cortex®-M33 and DSP Cores
- i.MX-RT1050: i.MX RT1050 Crossover MCU with Arm® Cortex®-M7 Core
- i.MX-RT1060: i.MX RT1060: Crossover MCU with Arm® Cortex®-M7
- i.MX-RT1064: i.MX RT1064: Crossover MCU with Arm® Cortex®-M7
- i.MX-RT1170: i.MX RT1170: 1 GHz Crossover MCU with Arm® Cortex® Cores
- i.MX-RT1160: i.MX RT1160 Crossover MCU Dual-Core Arm® Cortex®-M7 and Cortex-M4