DMR News

Advancing Digital Conversations

Arm integrates AI capabilities into IoT chips designed for edge computing applications.

ByYasmeeta Oon

Apr 27, 2024

Arm integrates AI capabilities into IoT chips designed for edge computing applications.

In an ambitious move to redefine the landscape of edge computing, Arm, the semiconductor architecture titan headquartered in Cambridge, England, has announced the launch of groundbreaking edge AI technologies. The unveiling of the Ethos-U85 Neural Processing Unit (NPU) and the Corstone-320 IoT Reference Design Platform marks Arm’s foray into integrating artificial intelligence capabilities directly into the fabric of edge computing devices. This innovation is poised to enhance the intelligence of the Internet of Things (IoT), making devices smarter, more efficient, and more autonomous.

At the heart of today’s announcement is the introduction of two pivotal technologies: the Ethos-U85 NPU and the Corstone-320 IoT Reference Design Platform. These developments by Arm are set to revolutionize edge AI applications through improved performance, energy efficiency, and ease of integration.

The Ethos-U85 represents the zenith of Arm’s NPU technology, offering unprecedented efficiency and power for edge AI applications. It marks a significant leap forward, boasting a performance increase of four times and a 20% uplift in energy efficiency over its predecessors. Designed to meet the rigorous demands of applications such as factory automation and smart home cameras, the Ethos-U85 stands out for its:

  • Scalable Architecture: With a range from 128 to 2048 MAC units, equating to 4 TOPs @1GHz, the Ethos-U85 is tailored to deliver top-tier performance while preserving energy efficiency.
  • Framework Compatibility: Support for popular AI frameworks like TensorFlow Lite and PyTorch, facilitating seamless integration for developers.
  • Versatile AI Use Cases: Compatibility with Transformer Networks and Convolutional Neural Networks (CNNs), enabling a wide range of AI tasks including vision and generative AI.

The Ethos family has already seen significant interest from the industry, with over 20 partners licensing the technology. Early adopters such as Alif Semiconductor and Infineon have recognized the Ethos-U85’s potential to spearhead innovation in edge AI solutions.

Accompanying the launch of the Ethos-U85 is the Corstone-320, a comprehensive platform aimed at expediting the development of edge AI systems. This platform melds the capabilities of the Arm Cortex-M85 CPU, Mali-C55 Image Signal Processor, and Ethos-U85 NPU to facilitate real-time processing of complex data types. Key features of the Corstone-320 include:

  • Integrated Hardware and Software: Offers a unified platform for rapid prototyping and development of edge AI applications.
  • Streamlined Deployment: Provides tools, support, and Arm Virtual Hardware to ease the deployment process, significantly cutting down the time-to-market for edge AI devices.

Arm’s architecture forms the backbone for chips used globally by millions of developers. With the release of these advanced designs, Arm’s licensed partners have the blueprint to design and manufacture their own innovative chips.

The introduction of the Ethos-U85 NPU and Corstone-320 platform signifies a monumental step in the evolution of edge computing. These technologies offer a host of benefits, including but not limited to:

  • Enhanced Performance: Devices powered by these technologies will exhibit significantly improved computation capabilities, enabling more complex and intelligent functionalities.
  • Greater Efficiency: The focus on power efficiency ensures that devices can deliver superior performance without sacrificing battery life or requiring extensive cooling solutions.
  • Seamless Integration: With support for leading AI frameworks and a range of AI tasks, developers can easily incorporate these technologies into their products, fostering innovation.

Leaders from Alif Semiconductor and Infineon have lauded Arm’s advancements, highlighting the transformative potential of the Ethos-U85 and Corstone-320 in edge AI applications. Reza Kazerounian, President of Alif Semiconductor, emphasized the importance of performance and efficiency for the next generation of edge AI applications. Similarly, Steve Tateosian, Senior Vice President at Infineon, acknowledged the sophisticated requirements of modern edge AI use cases and the role of Arm’s technologies in meeting these demands.

Arm’s latest venture into edge AI with the Ethos-U85 NPU and Corstone-320 platform is more than just a technological advancement; it’s a strategic move that positions Arm at the forefront of the AI revolution in edge computing. By providing the foundational technologies that combine high performance, efficiency, and ease of integration, Arm is not only redefining the capabilities of edge devices but also shaping the future of IoT and smart technologies. As the digital world continues to evolve, Arm’s innovations stand as a testament to the potential of integrating AI into the very core of edge computing, promising a smarter, more efficient, and interconnected future.

Quick Comparison: Ethos-U85 vs. Previous Generations
FeatureEthos-U85Previous Generation
Performance Boost4x
Power Efficiency20% higher
ArchitectureScalable from 128 to 2048 MAC unitsFixed architecture
AI FrameworksTensorFlow Lite, PyTorchLimited support
Use CasesVision, Generative AI, and morePrimarily vision tasks
Key Highlights of Arm’s Announcement
  • Ethos-U85 NPU: Unprecedented efficiency and power for edge AI, suitable for a wide range of applications.
  • Corstone-320 Platform: A comprehensive solution for accelerated development of edge AI systems, blending hardware and software for rapid prototyping.
  • Industry Adoption: Over 20 partners have licensed the Ethos technology, with early adopters showcasing potential for innovation in edge AI.
  • Strategic Implications: Arm’s advancements signal a significant shift towards integrating AI capabilities directly into edge computing devices, promising to transform the IoT landscape.

Related News:


Featured Image courtesy of DALL-E by ChatGPT

Yasmeeta Oon

Just a girl trying to break into the world of journalism, constantly on the hunt for the next big story to share.

Leave a Reply

Your email address will not be published. Required fields are marked *