NXP Semiconductors has expanded its eIQ AI and machine learning development software with two new tools designed to enhance edge AI deployment capabilities across their processor range. The first addition, eIQ Time Series Studio, introduces an automated machine learning workflow specifically engineered for time series-based machine learning models on MCU-class devices, including the MCX and i.MX RT crossover MCU portfolios.
The eIQ Time Series Studio supports a comprehensive range of input signals, including voltage, current, temperature, vibration, pressure, sound, and time of flight measurements. The platform enables multi-modal sensor fusion and features automatic machine-learning capabilities that can transform raw time-sequential data into actionable insights. The development environment includes tools for data curation, visualization, analysis, model auto-generation, optimization, emulation, and deployment. The system is designed to create optimized anomaly detection, classification, and regression libraries without requiring extensive data science or AI expertise.
The second major addition is the Large Language Model (LLM) Flow, which provides the fundamental architecture for implementing generative AI solutions at the edge. This tool is optimized for use with MPUs, particularly NXP’s i.MX family of applications processors. The LLM Flow incorporates Retrieval Augmented Generation (RAG) capabilities, allowing for secure fine-tuning of models using domain-specific knowledge and private data without compromising sensitive information security.
The expanded toolkit is designed to facilitate AI deployment across the full spectrum of edge processors, from small microcontrollers to powerful applications processors like the i.MX 95. This comprehensive approach enables reduced latency, enhanced user privacy, and lower energy consumption in edge AI applications. The system supports various model types, including generative AI, time series-based models, and vision-based models, providing developers with extensive flexibility in implementing AI solutions across different edge processing platforms.
These enhancements to the eIQ toolkit represent a significant advancement in making edge AI deployment more accessible and efficient, particularly for developers working across diverse markets and processing requirements. The platform’s automated workflows and optimization capabilities streamline the development process while maintaining performance, memory efficiency, and accuracy requirements for edge-based AI applications.
Leave a Reply