Arm adds Cortex-M42 to bring AI processing to small, cost-sensitive devices
The processor targets wearables, sensors and industrial controllers.
Loading page…
The processor targets wearables, sensors and industrial controllers.
Listen to this story
Arm’s Cortex-M42 gives chip designers a compact option for adding more machine-learning and signal-processing capability to microcontrollers used in wearables, sensors and industrial equipment. Arm says it delivers up to three times Cortex-M33’s vector compute capability and twice its vector power efficiency, but those figures apply to vector processing—not whole-device speed or battery life. The practical bet is that designers can handle tasks such as keyword spotting and equipment monitoring locally, while keeping the processor out of safety-critical motion control.
Cortex-M42 uses Armv8.1-M, Helium technology and support for Arm Custom Instructions, allowing chip partners to tailor it to workloads.
Arm added Cortex-M42 to Flexible Access and Total Access; as of November 2026, the programs had 319 members and 60 partners, respectively.
Arm reported more than 500 unique tape-outs through Flexible Access and more than 600 projects supported by Total Access since launch.
Arm is giving chip designers another way to add AI processing to devices with little room or budget for more computing. On November 10, 2026, it introduced Cortex-M42 and added the processor to Arm Flexible Access and Arm Total Access, targeting applications such as wearables, connected sensors and industrial controllers.
Cortex-M42 combines Arm’s Helium technology with a compact design based on AHB, a hardware interface. Arm says that combination makes it easier for silicon partners to upgrade microcontrollers based on Cortex-M4 and Cortex-M33 while adding more advanced machine learning and digital signal processing—the work of processing inputs such as audio and sensor signals.
The processor uses the Armv8.1-M architecture and supports Arm Custom Instructions. Those instructions give partners another way to optimize the processor for particular workloads.
Arm compares the new processor with Cortex-M33 on two measures: vector compute capability and vector power efficiency. The figures below are company claims about those measures, not a promise that an entire device will run three times faster or last twice as long on a battery.
Arm positions Cortex-M42 for applications where both cost and the space occupied by the processor matter. Its examples focus on interpreting signals close to the device collecting them, rather than assigning the processor every part of a machine’s operation.
Robotics provides a useful boundary for that pitch. Arm says Cortex-M42 could process signals locally and monitor the condition of motors and joints, helping identify unusual behavior or equipment wear. It explicitly separates that example from safety-critical motion control: the processor would monitor the machinery, not perform those safety-critical movement functions.
Arm Flexible Access offers low- or no-cost access to a portfolio of processor designs, development tools and technical resources.
Arm Total Access serves established semiconductor companies, equipment manufacturers, design manufacturers and system integrators as they develop, commercialize and scale products.
Arm reported 319 active Flexible Access members and more than 500 unique tape-outs as of November 2026. Total Access had 60 partners and had supported more than 600 projects since launch.
Loading discussion...
Join the conversation
Explain where you would draw the line.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.