Arm announces new chip architecture with focus on security and AI

ARMv9 features Confidential Compute Architecture and 'Realms' to shield sensitive data and code from the OS and non-permitted apps

Arm Ltd, the British chip architecture firm that licenses its designs to others, has unveiled a new generation of chip architecture, which the firm hopes, will be used in over 300 billion chips in the next 10 years.

Called V9 or ARMv9, the new technology is focused on three primary areas: security, artificial intelligence (AI), and overall computing power.

Arm's last major update to its instruction set architecture (ISA) was V8 that was unveiled in 2011 with the 64-bit AArch64 instruction set. V8 enabled Arm devices to make the leap from low-power mobile devices to ultra-fast supercomputers.

Arm is continuing to use AArch64 as the baseline instruction set for ARMv9 architecture, although the firm has expanded it with a many new features, all aimed at improving security and performance and taking over data centres and AI applications.

Some of the new features in ARMv9 architecture include SVE2 (the latest version of the firm's Scalable Vector Extensions (SVE) technology), Realm Management Extension (RME), BRBE (Branch Record Buffer Extension), Embedded Trace Extension (ETE), Trace Buffer Extension (TRBE) and Transactional Memory Extension (TME).

ARMv9 includes Arm's new Confidential Compute Architecture (CCA). With V9, Arm is also introducing a new 'Realms' concept, which, the firm says, will enable developers to create applications where the sensitive data and code is shielded from the OS and other apps on the device, by executing them in a different environment.

"We see no reason why the Realm technology cannot scale up to provide security in the most high-performance systems that are out there," Arm CEO Simon Segars said during a news briefing.

The company says it has also added features for AI chips for tasks like image and speech recognition. It also claims that the new SVE2 feature will give ARMv9 chips an edge when processing machine learning, 5G, and augmented and virtual reality workloads locally on their CPU.

According to Arm, the new architecture will enable next generations of mobile and cloud chips to have performance boosts of more than 30 per cent.

The new V9 chip architecture has come at the time when Arm is preparing for its planned $40 billion (£29 billion) merger with American graphics hardware giant Nvidia Corp.

Nvidia expects the deal to help expand its presence in AI and mobile computing, enabling it to reach more devices and platforms, such as PCs, smartphones and autonomous vehicles.

Commenting on Arm's new chip architecture, Brian Kelleher, senior vice president of hardware engineering at Nvidia, said that their company "sees enormous opportunities to bring the transformative powers of AI deeper into gaming, autonomous vehicles, enterprise data centres and embedded devices".

"Through our ongoing collaboration with Arm, we look forward to using Armv9 to deliver a wide range of once unimaginable computing possibilities."

Min Goo Kim, executive vice president of SoC development at Samsung Electronics Samsung Electronics, said: "Arm's next-generation Armv9 architecture offers a substantial improvement in security and machine learning, the two areas that will be further emphasized in tomorrow's mobile communications devices."

"As we work together with Arm, we expect to see the new architecture usher in a wider range of innovations to the next generation of Samsung's Exynos mobile processors."