Key insights (AI-powered):
Nordic’s move effectively breaks down the stack between ultra-low power silicon connectivity, ML tools and lifecycle cloud services.
By embedding an NPU and TinyML models into a mainstream wireless SoC, Edge AI moves from a niche add-on to a standard feature in tracking, wearables and industrial sensing, which will pressure competitors to integrate similar accelerators.
The integrated Edge AI Lab and OTA pipeline lowers the hurdles to AI deployment for traditional embedded teams and accelerates mass experimentation with on-device analytics.
This marks a broader IoT shift from simple collection and reporting to continuous, local decision making at the endpoint.
The new nRF54L series SoC with NPU and Nordic Edge AI Lab makes on-device intelligence easily accessible and extremely power efficient
Nordic Semiconductor, a global leader in low-power wireless connectivity solutions, brings AI intelligence and functionality to the smallest battery-powered IoT devices. With an industry-leading, ultra-low-power Edge AI solution, Nordic is accelerating the introduction of a new generation of devices with integrated Edge AI intelligence – combining power efficiency with unmatched ease of use for developers.
Nordic Semiconductor simplifies edge AI for billions of IoT devices
- nRF54LM20B– New ultra-low power wireless SoC with large memory and Axon NPU
- Custom Neuton Models – Industry-leading, ultra-small edge AI models
- Nordic Edge AI Lab– Development tool that simplifies and accelerates edge AI development
“AI factories train intelligence, but Nordic deploys it – on the device, at the edge, where the world happens.” – said Vegard Wollan, CEO of Nordic Semiconductor.
“Edge AI is no longer optional – it is the only way to deliver security, privacy and sustainability at scale. Nordic’s Edge AI solution enables millisecond decisions with no round-trip latency to the cloud, ensures compliance through local processing, and delivers radically improved battery life for billions of connected devices. This is the new standard for ultra-low power edge AI, and Nordic is defining it.”
Next generation nRF54LM20B with Axon NPU
In 2023, Nordic acquired Atlazo and its Axon technology. The nRF54LM20B SoC is the first member of the nRF54L series with large memory and integrates the Axon Neural Processing Unit (NPU), a highly efficient AI hardware accelerator, to accelerate demanding edge AI workloads. Axon delivers up to 7x faster performance and up to 8x higher power efficiency compared to competing solutions for tasks such as sound classification, keyword recognition and image-based recognition.
The nRF54LM20B SoC combines the Axon NPU with 2MB NVM, 512KB RAM, a 128MHz Arm® Cortex-M33 plus RISC-V coprocessor, high-speed USB, up to 66 GPIOs, and Nordic’s fourth-generation 2.4GHz ultra-low-power radio that supports Bluetooth® LE, Bluetooth Channel Sounding, Matter over Thread, and more.
Industry-leading ultra-small Neuton Edge AI models
The Neuton models are ultra-small, CPU-driven edge AI models that are typically less than 5 KB in size and are up to 10 times smaller, faster and more efficient than other CPU-driven models. Nordic Edge AI Lab helps developers build custom Neuton models for anomaly detection, activity and gesture recognition, biometric monitoring, and more – delivering privacy-preserving, real-time information on tiny batteries and limited storage, without cloud dependency.
Edge AI will become a natural part of any new product design
With Nordic Edge AI Lab and Neuton models, intelligence goes from concept to reality without complexity. A recent deployment illustrates this: a global supply chain solution effortlessly upgraded its intelligent tracking devices with AI models built in the Nordic Edge AI Lab, enabling the detection of real-world handling events such as shock, vibration and transport directly on an nRF54L series SoC. Thanks to Nordic’s nRF cloud lifecycle services, these AI-driven insights were able to be rolled out across the entire fleet without disruption to operations.
“With Nordic Edge AI Lab, Neuton models and the Axon NPU, Nordic makes advanced on-device AI practical for every embedded developer.” said Oyvind Strom, EVP Short-Range BU at Nordic Semiconductor. “Developers get the simplicity to act quickly and the breakthrough performance to scale from wearables to industrial sensors – all enabled by Nordic’s proven ultra-low power hardware solutions.”
As intelligence continues to shift, the need for OTA management and deep observability grows. At the same time, cloud-based lifecycle services remain critical for device management, embedded observability and location services. Manufacturers increasingly need continuous insight into device performance – not only to improve products in real time, but also to meet growing regulatory and customer demands. Using data from provisioned devices improves features and optimizes performance without impacting the user experience. This ensures that connected products can evolve safely and efficiently throughout their entire life cycle.
Availability
Nordic Edge AI Lab and custom Neuton models are now available for Nordic Wireless nRF54 series SoCs and cellular IoT SiP modules. The nRF54LM20B SoC with Axon NPU is currently being sampled to select customers. Wide availability for development is expected early in the second quarter of 2026.
