u-blox has introduced ALMA-B2, a Bluetooth LE module series built on Nordic Semiconductor’s nRF54LM20 platform and aimed at compact IoT products that need wireless connectivity and edge processing in the same design. The series brings Bluetooth LE 6.0, Thread and Zigbee support, Channel Sounding, security features, and antenna options into a certified module format.
ALMA-B2 is used to add short-range wireless communication and local processing to low-power connected equipment. The ALMA-B21 version goes further by adding an Axon Neural Processing Unit, giving developers a hardware path for Edge ML workloads rather than running those algorithms only on the MCU.
The module is likely to be used in products such as industrial sensors, asset trackers, healthcare devices, robotics equipment, and smart home hardware, where small changes in response time and battery use can have a real impact.
Bluetooth Channel Sounding For Distance-Aware Devices
Bluetooth LE 6.0 support is joined by 802.15.4 connectivity, covering Thread and Zigbee designs as well as standard Bluetooth LE products. Channel Sounding is the more specific addition for designs that need distance measurement between devices. That gives ALMA-B2 a route into proximity sensing, asset tracking, access systems, and other equipment where the radio link is doing more than moving small packets of data.
The module will be available with embedded and external antenna options. That gives developers a practical choice between a simpler certified radio design and layouts where antenna placement needs to be handled separately because of the enclosure, board shape, or RF environment.
Nordic nRF54LM20 Platform And Edge ML Hardware
Inside the module is a 128 MHz Arm Cortex-M33 processor together with 2 MB of non-volatile memory and 512 kB of RAM. A RISC-V coprocessor is also included as part of the platform architecture. ALMA-B21 variants add an Axon NPU for hardware-based machine learning acceleration. According to u-blox, the dedicated NPU can deliver up to 15x higher performance compared with running ML algorithms directly on the MCU alone while also reducing power consumption.
The wider shift toward Edge ML is becoming easier to spot across embedded hardware. Instead of constantly moving sensor data back to cloud infrastructure, more systems are now expected to process at least part of that workload locally. In industrial monitoring equipment or healthcare wearables, that can reduce latency while lowering wireless transmission overhead.
Security Features And Embedded Interfaces
The ALMA-B2 series includes Arm TrustZone support, cryptographic accelerators, tamper detection, immutable boot partition support, and debug port protection. Those features are increasingly becoming baseline requirements for connected products expected to remain deployed for years.
HS-USB support is also included for embedded systems that need higher throughput than standard peripheral interfaces typically provide.
u-blox says the modules are designed for operation from -40 °C to 85 °C and will be offered in compact formats with both embedded and external antenna options. Early samples are expected to become available in the coming weeks.
Learn more and read the original announcement at www.u-blox.com
Technology Overview
The u-blox ALMA-B2 series is a compact wireless module family based on Nordic Semiconductor’s nRF54LM20 platform. The modules support Bluetooth LE 6.0, Thread, Zigbee, and Bluetooth Channel Sounding, while ALMA-B21 variants add a dedicated Axon NPU for hardware-based Edge ML acceleration. Integrated hardware includes a 128 MHz Arm Cortex-M33 processor, 2 MB non-volatile memory, 512 kB RAM, and a RISC-V coprocessor.
Frequently Asked Questions
What is the u-blox ALMA-B2 module used for?
The ALMA-B2 series is intended for low-power wireless products that need local processing and secure short-range connectivity. Example applications include industrial sensors, healthcare equipment, robotics systems, smart home products, and asset trackers.
Does the ALMA-B2 support Edge ML workloads?
Yes. ALMA-B21 variants include a dedicated Axon Neural Processing Unit for hardware-based Edge ML acceleration. Which wireless protocols are supported by ALMA-B2? The modules support Bluetooth LE 6.0 together with 802.15.4 technologies including Thread and Zigbee.
You may also like
