Introducing new Arm Roadmap Guarantee and other Arm IoT initiatives for accelerating Endpoint AI

搜索最新SoC解决方案
关于汽车电子...物联网...安全...音频...视频

By Mohamed Awad, vice president, IoT business at Arm

November 18, 2020 -- While the Internet of Things (IoT) revolution may still be in its early days, there is already one undeniable similarity to the smartphone revolution - the Arm ecosystem is the driving force behind both. Which is why it’s an incredible opportunity to step into the role of leading Arm’s IoT business and work with our ecosystem to continue harnessing and unleashing the potential of this market.

Arm is in a unique position to drive the innovation and adoption of IoT everywhere, and here are three numbers that highlight why we are unique – 70, 100, and 180. Arm technology touches 70% of the world’s population, thanks to our partners having shipped 180 billion Arm-based chips since Arm was founded 30 years ago. And if you look a little closer at those 180 billion chips, more than 100 billion have shipped in just the past five years alone.

But it’s what is beyond the numbers that counts. Arm’s technology is the foundation upon which diverse, innovative, and life-changing applications are built, from smart healthcare to remote monitoring, from industrial robotics to the connected home.

If we are going to truly rival the smartphone ecosystem in economic value, there is still more to do in enabling the development, deployment, and monetization of IoT devices. To that point, here are some important updates we’re announcing today:

• Expanding the endpoint AI capabilities of the Arm Flexible Access portfolio with the inclusion of the Arm Cortex-M55 and Ethos-U55
• A new Roadmap Guarantee for our Arm Flexible Access customers
• Focusing our software investment on simplifying development to unleash the potential of our diverse IoT ecosystem

New additions to Arm Flexible Access: Putting endpoint AI in everyone’s hands

We expect endpoint AI (IoT endpoint devices locally processing data into actionable insights) will accelerate the value generated from IoT and drive adoption. To do this at a cost point and power budget that enables mass deployment of efficient and intelligent endpoint AI, earlier this year we introduced the industry’s first microNPU, the Ethos-U55, and the Cortex-M55 microcontroller. And most recently, we announced Ethos-U65, which brings the same capability to higher performance points and extends microNPU applicability to Cortex-A based systems.

To help further accelerate endpoint AI deployments, the Flexible Access portfolio has been expanded to include both Cortex-M55 and Ethos-U55, giving Flexible Access partners the freedom to explore, experiment and design with these products without any up-front license costs.

A new Roadmap Guarantee for Flexible Access partners

Since the launch of Flexible Access in summer 2019, more than 100 innovative companies, from startups, to ASIC houses to established semiconductor companies, have signed on. What’s particularly exciting is that more than half of these are first-time Arm silicon partners.

As more partners continue to sign on, we’re evolving the program to ensure they have the confidence to develop the long-term roadmaps their IoT solutions will require by introducing the Roadmap Guarantee for Flexible Access. The Roadmap Guarantee ensures:

• Arm will continue to offer the CPUs included in the Flexible Access portfolio for five years.
• A commitment that Arm will add future Cortex-M and Ethos-U products to Flexible Access shortly after they come to market.

Keeping it simple and collaborating in a software-defined world

The updates to Flexible Access give our silicon partners greater access to Arm endpoint AI IP, but also benefit the software developer ecosystem. We have never been more committed to making the software development experience on Arm as seamless and simple as possible. The service-orientated architecture of the cloud native datacenter is colliding with the weird, wonderful, and diverse hardware that makes up embedded edge and endpoint devices. Working with our ecosystem we are focused on ensuring all of the varied and evolving workloads are portable, maintainable, and easily deployed across devices and operating systems while seamlessly connecting to any cloud service.

This is why efforts like Project Cassini, an open, standards-based initiative to deliver a cloud-native software experience across a secure Arm edge ecosystem are so critical. As well as expanding our collaborations with partners like Microsoft to deliver a unified and streamlined AI experience for developers to deploy a common, integrated Azure-based toolchain targeting Arm-based silicon.

While Arm continues to invest in IoT developer enablement, the amount of work already underway by our silicon partner ecosystem is extraordinary. For example, Renesas recently announced they are expanding microcontroller (based on Cortex-M) support for Google’s TensorFlow lite to enable tinyML powered solutions for applications such as motor control and predictive maintenance for smart homes and industrial automation. Watch this space for more news from our partner ecosystem.

Looking ahead

Arm technology enabled a smartphone revolution that changed the way we used technology forever, but we need more from the IoT revolution. We need IoT to make our lives better across everything we do. We continue to set a high bar with smartphone innovation, but for IoT innovation and its unlimited potential to change lives, there really is no high bar or ceiling for what the Arm ecosystem can achieve together.

About Arm

Arm technology is at the heart of a computing and data revolution that is transforming the way people live and businesses operate. Our energy-efficient processor designs and software platforms have enabled advanced computing in more than 180 billion chips and our technologies securely power products from the sensor to the smartphone and the supercomputer. Together with 1,000+ technology partners we are at the forefront of designing, securing and managing all areas of AI-enhanced connected compute from the chip to the cloud.

 Back