100 billion devices ready for AI by 2025, says Arm CEO

His presentation focused on how Arm’s technology accelerates time-to-market with efficient solutions, highlighting the company’s impact on various industries and everyday life.

The role of Arm in the AI revolution

Rene Haas began by reflecting on Arm’s longstanding influence within the tech ecosystem. He noted that Arm has collaborated with over 18 million software developers and has shipped nearly 300 billion ARM-based chips. This extensive reach underpins the company's AI and computing advancements.

He remarked: “Arm's high-performance, power-efficient processors enable fast computing in the most complex AI supercomputers to low-power Edge devices.”

Haas highlighted the transformative impact of AI on everyday activities. He outlined how AI is improving work, entertainment, connectivity, and overall wellbeing. Arm's technologies are designed to support this transformation by providing robust AI computing platforms that are adaptable across various devices and applications.

He stated: “The future of AI is built on Arm, changing lives for the better.”

Addressing energy efficiency

With the growing energy demands of AI computing, Haas underscored the importance of power efficiency in sustaining this growth. He presented data projecting that by the end of the decade, 20% of the US energy grid could be consumed by data centres if current trends continue. He stressed: “There is no way that this can continue without some significant changes. Arm's energy-efficient designs are crucial to managing this demand sustainably.”

Haas discussed Arm's strategic partnerships with tech giants such as Samsung, Google, and Meta. These collaborations are aimed at optimising AI models and integrating them across different devices and platforms.

He mentioned: “Our collaboration with industry leaders ensures that AI runs efficiently across a wide range of devices, from the Cloud to the Edge.”

Haas elaborated on how these partnerships enhance the AI ecosystem, making it easier for developers to create and deploy AI applications effectively.

Software ecosystem

A significant part of Haas’s speech focused on Arm's software ecosystem, which supports all major operating systems and boasts over 80 million developers. This ecosystem is vital for driving innovation and ensuring that AI applications can run seamlessly on various hardware platforms.

Haas explained: “Arm's software ecosystem is unparalleled, providing a robust foundation for developers to build applications that leverage the full potential of our hardware.”

New product announcements

Chris Bergey, Senior Vice President and General Manager, Client Line of Business introduced new products, including the Neoverse Compute Subsystems (CSS) for Client, which features high-performance, power-efficient CPUs and GPUs. These innovations are designed to accelerate AI workloads and offer scalable solutions across different computing environments.

CSS for Client includes the latest Armv9.2 Cortex CPU cluster and Arm Immortalis and Mali GPUs, CoreLink Interconnect system IP, and production-ready physical implementations for the CPUs and GPUs on the 3nm process on leading foundries. The platform provides the fastest path to production silicon for our partners. The physical implementations allow Arm’s partners to unlock all the benefits of the leading-edge 3nm process, while enabling highly flexible, customisable silicon designs.

Bergey highlighted: “Our new CPUs and GPUs are built to deliver exceptional performance and energy efficiency, enabling faster and more effective AI computing.”

Arm also recently launched Kleidi, a comprehensive programme of software and community engagements aimed at accelerating AI. The first component is the Arm Kleidi libraries for popular AI frameworks, enabling developers to seamlessly access the AI capabilities of the widely used Arm CPU, which handles most of the world’s AI inference workloads from Cloud to Edge. Developers can leverage over 20 years of Arm architectural innovation, which has consistently improved AI capabilities and performance – from the Armv7 architecture that introduced the Advanced Single Instruction Multiple Data (SIMD) Extension for machine learning (ML) workloads to the Armv9 architecture that accelerates and protects advanced generative AI workloads on the Arm CPU.

KleidiAI, designed for all AI workloads, and KleidiCV, optimised for computer vision (CV) workloads on Arm CPUs across all tiers, are embedded directly into popular AI frameworks, requiring no action from developers. This enables developers to effortlessly utilise the AI capabilities of the Arm CPU, allowing them to quickly build AI-based applications with the highest possible performance across a broad range of devices.

AI devices milestone

Haas made a significant projection regarding the future of AI devices, stating: “We anticipate that by the end of 2025, there will be 100 billion ARM devices ready for AI.”

This milestone underscores Arm’s influence and the widespread adoption of its technologies in the AI landscape.

Concluding his keynote, Haas emphasised Arm's commitment to advancing AI technology and supporting the global AI community. He expressed confidence in Arm's ability to lead the industry through continuous innovation and strategic collaboration. He stated: “We are dedicated to driving the AI transformation with our comprehensive computing platform. The future of AI is incredibly bright, and Arm is at the heart of this exciting journey.”