Arm unveils new platform naming architecture to support AI evolution

The move forms part of Arm’s broader strategy to support performance and efficiency at the system level as AI workloads reshape infrastructure, automotive, client, mobile, and IoT computing.

Chief Executive Officer Rene Haas said the industry faced a core challenge: boosting performance without escalating power consumption. As AI increasingly drives demand for more compute across all devices – from data centres to earbuds – the need for efficient, scalable platforms has intensified.

To address this, Arm expanded its Compute Subsystems (CSS) offering across key segments including infrastructure, client, automotive, and the Edge AI platform for IoT. The company positioned CSS as a critical enabler for performance-per-watt gains, faster time-to-market, and system-level integration.

As part of this platform-first approach, Arm introduced a streamlined naming convention that clearly identifies each platform’s end market:

• Arm Neoverse for infrastructure
• Arm Niva for PC
• Arm Lumex for mobile
• Arm Zena for automotive
• Arm Orbis for IoT

The Mali brand will remain as Arm’s GPU IP offering, integrated within these platform solutions. Additionally, IP numbering will be simplified and aligned with platform generations. New labels – Ultra, Premium, Pro, Nano, and Pico – will indicate performance tiers to ease navigation for customers and developers.

Arm stated that the new structure reflects growing adoption of its platform at the system level, rather than at the individual IP level. It aims to reduce complexity for partners while increasing confidence and accelerating development cycles, particularly in AI-driven environments.

Haas concluded that the changes signal a future where AI compute is “built on Arm.”