Arm announced an expansion of its Flexible Access programme to include its Armv9 Edge AI Platform, aiming to simplify access to advanced compute for companies developing next-generation Edge and IoT devices.
The move reflected Arm’s continued mission to make its technology more accessible across its ecosystem, from small startups to global OEMs, and to lower barriers to entry for AI-capable hardware development.
The Armv9 Edge AI Platform, introduced earlier this year, integrated the Arm Cortex-A320 CPU and Arm Ethos-U85 NPU. Together, these enabled on-device AI models with more than one billion parameters and supported a range of applications, including smart cameras, connected home systems, robotics, and industrial automation.
The platform also incorporated Armv9 security features, such as Pointer Authentication Code (PAC), Branch Target Identification (BTI), and Memory Tagging Extension (MTE), to protect data and applications at the Edge. Scalable Vector Extension 2 (SVE2) enhanced ML performance, while the Ethos-U85 NPU provided efficiency for AI workloads, creating a strong foundation for devices that processed data locally and responded intelligently through vision, voice, and gesture recognition.
Cortex-A320 would be available through Arm Flexible Access in November 2025, followed by Ethos-U85 in early 2026.
Arm stated that this development represented a significant milestone in its commitment to Edge AI innovation. Research by VDC suggested that by 2028, AI would become the dominant technology used across IoT projects, positioning Arm’s platform at the centre of that evolution.
Since its launch, Arm Flexible Access had enabled partners to experiment with Arm IP under a low-cost or no-cost model, only paying licence fees for technology included in final designs. The initiative had supported over 300 active members and resulted in around 400 successful tape-outs over five years. More than half of these designs targeted embedded and IoT markets, enabling intelligent functionality closer to where data was generated.
Companies such as Raspberry Pi, Weeteq, Hailo, and SiMa.ai had already leveraged Arm Flexible Access to accelerate innovation in Edge AI.
Arm stated that extending access to its Armv9 architecture would give the next generation of developers the performance, security, and efficiency required to meet growing demand for intelligent Edge systems.
Paul Williamson, SVP and GM of the IoT Business, Arm commented: “Our goal has always been to democratise access to world-class technology. With Armv9 now part of Flexible Access, we are empowering innovators to build smarter, more secure, and more efficient Edge AI devices that bring intelligence closer to where data is created.”
