New Arm 'AI Anywhere' Chips Boost IoT, Edge Performance
- By John K. Waters
Chip designer Arm just announced two new processors with the potential to push artificial intelligence (AI) and machine learning (ML) into billions of small, power-and-memory-constrained devices.
The new additions to the ARM AI platform -- the Cortex-M55 and the Ethos-U55 -- deliver a combined 480x increase in ML performance to microcontrollers, the company said.
The Cortex-M55 is "the most AI-capable" of the company's Cortex-M processor designs. It's also the first based on the Arm8.1-M architecture. Arm is billing its new Ethos-U55 as the industry's first micro neural processing unit (NPU). Both deliver "unprecedented levels of on-device ML processing," the company said. The NPU utilizes advanced compression techniques to lower power requirements and ML model sizes to enable execution of neural networks that previously ran only on larger systems.
"Enabling AI everywhere requires device makers and developers to deliver machine learning locally on billions, and ultimately trillions of devices," said Dipti Vachani, senior vice president and general manager of Arm's Automotive and IoT line of business. "With these additions to our AI platform, no device is left behind, as on-device ML on the tiniest devices will be the new normal, unleashing the potential of AI securely across a vast range of life-changing applications."
Arm Holdings, which is owned by the Japanese SoftBank Group, is a global provider of so-called "silicon IP for devices." It doesn't make chips; it designs them, manages the instruction set, and creates new versions of the core architecture. It licenses all that to silicon designers and device manufacturers. Those companies are free to improve on the design and pair it with their hardware of choice.
"As the IoT intersects with AI advancements and the rollout of 5G, more on-device intelligence means that smaller, cost-sensitive devices can be smarter and more capable, while benefiting from greater privacy and reliability due to less reliance on the cloud or internet," the company said in its announcement. "By delivering this intelligence on microcontrollers designed securely from the ground up, Arm is reducing silicon and development costs and speeding up time to market for product manufacturers looking to efficiently enhance digital signal processing (DSP) and ML capabilities on-device."
Software development in the AI-for-IoT space is highly complex, Vachani noted in a blog post, spanning a diverse range of use cases and hardware. Arm sees making sure developers lives don't get any harder as they move through technology iterations as its primary imperative, he said. "The work we've done means any developer writing ML code for ARM Cortex-M based devices will be able to port that work to use on Cortex-M55 and Ethos-U55 based chips,
John K. Waters is the editor in chief of a number of Converge360.com sites, with a focus on high-end development, AI and future tech. He's been writing about cutting-edge technologies and culture of Silicon Valley for more than two decades, and he's written more than a dozen books. He also co-scripted the documentary film Silicon Valley: A 100 Year Renaissance, which aired on PBS. He can be reached at firstname.lastname@example.org.