#USA #Edge AI #Nuremberg #Nordic Semiconductor #nRF54L Series

Enhancing Edge AI Capabilities with Nordic Semiconductor

Nordic Semiconductor, a leader in low-power wireless connectivity solutions, is making headlines by enhancing its nRF54L series to empower developers in the field of Edge AI. As battery-powered devices proliferate, the importance of efficient computing solutions becomes paramount. The company's latest model, the nRF54LM20B System on Chip (SoC), introduces a Neural Processing Unit (NPU), making advanced AI capabilities accessible to a broader audience.

Unleashing Real-Time AI Performance

One of the standout features of the nRF54LM20B is its integrated NPU, designed to accelerate TensorFlow Lite models significantly. This SoC can perform tasks up to 15 times faster while consuming considerably less energy than traditional Arm Cortex processors. Importantly, it exhibits eight times greater energy efficiency compared to its closest competing Edge AI solutions. Such performance enhancements mean it can handle Edge AI workloads from high-throughput sensor data, audio inputs, and event-driven tasks seamlessly, all powered by a miniature battery.

According to Øyvind Strøm, VP of Marketing at Nordic Semiconductor, “This next-generation capability revolutionizes what small battery-operated devices can perceive and interpret in the real world. By integrating powerful intelligence directly into the component without compromising efficiency, latency, or system complexity, we are ushering in a new category of products that are significantly smarter and more responsive than their predecessors.”

Simplifying Edge AI Development

Nordic aims to simplify the development of Edge AI by combining advanced hardware acceleration with flexible model support and user-friendly development tools. This makes integrating AI into consumer products a straightforward process. The nRF54LM20B provides the necessary performance for real-time intelligence across various applications, including activity sensors, smart home devices, industrial and medical sensors, and tracking devices.

These devices are designed to facilitate precise activity detection, natural responses to sound or presence, instant anomaly identification, and motion interpretation without relying on cloud processing, which can be time-consuming and energy-intensive.

Streamlined Workflow for Developers

To further reduce barriers for developers, Nordic has enhanced its low-power Edge AI solutions, allowing for a smoother pathway to adding intelligence to a wide range of battery-powered IoT devices. Developers now can create and optimize Edge AI models directly on the device. They can customize parameters and specify the model architecture for either classification or regression tasks, allowing for the design of sophisticated Edge AI architectures tailored for wearable devices, industrial, and smart home sensors.

Moreover, generating a custom activation word has become incredibly straightforward. Developers can create unique voice recognition models from a simple text input, such as “Hello Nordic,” eliminating the need for thousands of audio samples which traditionally complicate the implementation of voice commands in smart speakers, activity sensors, and battery-operated audio accessories.

Expansive Tool Ecosystem

Models generated using Nordic's new online tool, Nordic Edge AI, are now optimized for swift inference on the NPU or efficient execution on the CPU. This flexibility addresses compatibility concerns between models, tools, and firmware, enabling functionality like audio classification, anomaly detection, advanced IMU processing, predictive maintenance, environmental monitoring, and signal processing.

Overall, these updates substantially reduce the development time typically required to integrate AI capabilities into embedded components, accelerating the journey from concept to prototyping.

Product Availability

The nRF54LM20 DK is now available through Nordic's distributor partners. For inquiries regarding samples of the nRF54LM20B SoC, please contact the Nordic sales team. Mass production is slated to commence in the second quarter of 2026.

Meet Nordic at Embedded World 2026

Join Nordic Semiconductor at Embedded World 2026 for live demonstrations and presentations at Hall 4A, Stand 310 from March 10-12, 2026. Experience firsthand the future of Edge AI capabilities for battery-powered devices and see how Nordic is leading the charge in this innovative space.