Unlock the Benefits of Artificial Intelligence for IoT Devices
The convergence of AI, 5G, and IoT technologies is resulting in a need for real-time information and responses. Network latency, data privacy, and consumer expectations around user experience and environmental impact are driving a shift to computing at the endpoint. Combining on-device intelligence with AI and ML capabilities has the potential to unlock new use cases and applications – we call this endpoint AI.
Arm offers new compute technologies coupled with software and tools to help companies streamline the design, development, and support for AI-based IoT applications. The Cortex-M55 processor and Ethos family of microNPUs, combined with Arm tools deliver up to a 480x leap in ML performance, bringing endpoint AI to billions of devices.
The Dawn of Endpoint AI: Bringing Compute Closer to Data
Only 1 percent to 12 percent of company data is actually analyzed, according to a new report by Strategy Analytics. The ability to perform advanced localized processing closer to where data is collected results in faster and more accurate responses, allowing companies to maximize data insights.
This new report takes an in-depth look at endpoint AI and the new opportunities it enables, and explores the potential benefits and impacts of implementing endpoint AI across key application areas and use cases.
Endpoint AI Applications
As AI mainstream popularity grows, so does the range of applications and opportunities businesses seek to develop, which span the breadth of IoT segments. Endpoint AI use cases align to three distinct areas:
Vision, such as object classification; Voice, including keyword detection; and Vibration, for example sensor fusion.
Intelligent Thermal Vision
Conservation challenges in remote areas require an energy-efficient, cost-sensitive solution. Arribada’s vision-based ML technology uses Arm microcontrollers and Mbed OS to identify and monitor wildlife.
Advanced Voice Control
Sensory’s TrulyHandsfree keyword recognition software works with Cortex-M55 to unlock high-accuracy, low-power, customizable voice control for devices and applications.
Cartesiam’s industry-agnostic AI solution enables anomaly detection and signal classification on microcontrollers based on Cortex-M, to monitor vibrations and other signal types and predict system failure.
Enabling Intelligence on the Smallest Devices
Advances in AI deployments mean that even the smallest, cost-sensitive devices can be made smarter and more capable. Arm IP, software, tools, and our partner ecosystem make development quicker and easier for product manufacturers implementing DSP and ML capabilities on-device.
Cortex-M55, Arm’s most AI capable Cortex-M processor, and Ethos microNPUs bring maximum performance, power and area efficiency for ML inference. Learn more about these best-in-class processors and how partner collaboration is pushing the boundaries of hardware design.Explore Cortex-M55
Software and Tools
Arm empowers billions of software developers with IoT and ML technologies, making it easier to innovate devices of the future. Arm’s platform for building and deploying endpoint AI devices enables software developers to unlock the full potential of Arm-based hardware by supporting any workload, device, software stack, and cloud.Learn More
Talk with an Expert
Find out how Arm’s new processors are enhancing on-device machine learning capabilities to help drive innovation and open up new business opportunities.
Endpoint AI Resources
In the Race to Realize the IoT There’s No Need to Run Alone
Arm and our world-leading ecosystem addresses the challenges of IoT development and deployment.
Cambridge Consultants: How We're Pushing the Endpoint AI Envelope
Unlocking the benefits of AI for IoT devices, balancing latency, privacy and power.
- TinyML Enables Smallest Endpoint AI Devices
- AI for IoT Devices
- Extending the Performance of Arm’s ML Portfolio for Endpoint Devices