eIQ software has been upgraded! Two new tools have been added to make edge AI deployment and use easier

NXP Semiconductors has announced the addition of two new tools to its eIQ AI and machine learning development software to make it easy to deploy and use AI edge on a variety of edge processors.

eIQ Time Series Studio provides an automated machine learning workflow that makes it easy to develop and deploy time series-based machine learning models on MCU-class silicon, such as the MCX series MCU portfolio or the i.MX RT crossover MCU portfolio .

GenAI Flow Provides building blocks for large language models (LLMs) that support generative AI solutions. These solutions are used in conjunction with MPUs, such as the NXP i.MX family of application processors , to simplify deployment of the intelligent edge by training the LLMs on contextual data. For example, an appliance equipped with an LLM trained on the user manual can communicate with the user in natural language, informing the user how to use a specific function, perform a specific task, or optimize usage and maintenance.

There are many benefits to deploying AI at the edge, including reduced latency, enhanced user privacy protection, and reduced energy consumption. The NXP eIQ toolkit extension significantly simplifies and accelerates the deployment process, giving developers access to a wider range of model types, including generative AI, time series-based models, and vision-based models. In addition, users can deploy models on a variety of edge processors.

“AI is key to enabling predictions and automation based on user needs, but it must be developed in a way that works for edge deployment,” said Charles Dachs, senior vice president and general manager of Industrial and IoT at NXP Semiconductors. “NXP provides ready-to-use tools for small AI models on MCUs such as the MCX series, crossover MCUs such as the i.MX RT700 , and large generative AI models running on more powerful devices such as the i.MX 95 application processors. This gives developers a rich choice of AI models and AI-enabled edge processors, making edge AI truly applicable to application developers across industries.”

eIQ Time Series Studio simplifies and accelerates the development and deployment of time series-based AI models, supporting multiple input signals (such as voltage, current, temperature, vibration, pressure, sound, flight time and signal combinations) and multimodal sensor fusion. With automatic machine learning capabilities, developers can extract meaningful insights from raw time series data and quickly build AI models that meet performance, memory, Flash storage size and accuracy requirements. The tool provides a complete development environment, including data management, visualization and analysis, as well as automatic model generation, optimization, simulation and deployment. The interface is simple and intuitive, and software developers do not need to have deep data science or AI expertise to create optimized anomaly detection, classification and regression libraries.

NXP's GenAI flow enables generative AI applications to be deployed on edge devices. This software flow provides methods to optimize generative models and provides retrieval-augmented generation (RAG), which can fine-tune models in a secure manner using domain-specific knowledge and private data without disclosing sensitive information to model or processor providers. By connecting multiple modules into a single flow, customers can easily customize LLMs for their tasks and optimize them for deployment at the edge using MPUs such as NXP's i.MX 95 application processors.