What Tech Innovations Did Arm Deliver in June 2025?

Technology innovations at Arm are accelerating across the compute continuum—reshaping industries, enabling breakthroughs, and redefining how developers build the future. 

At Arm, this progress is powered by a global ecosystem committed to making compute more scalable, efficient, and intelligent. June 2025 brought exciting developments in AI safety, mobile graphics, open source tools, and automotive software. 

From a new Arm naming architecture that simplifies platform alignment to breakthroughs in cloth simulation and safer LLM design, this roundup highlights key moments from across the Arm ecosystem—each one a step forward in transforming how we compute, connect, and create.  

What you need to know about Arm’s new product naming strategy

Arm has made a bold move to clarify and modernize how its products are understood across industries. In a strategic rebrand unveiled by CEO Rene Haas in May, Arm introduced a new naming architecture designed to better reflect its full-platform approach—from cloud infrastructure to automotive, mobile, and client devices.

To that effect, this Arm Community blog unpacks the technical rationale behind this shift and the evolution of Arm’s naming conventions. With distinctive new names like Neoverse, Niva, Lumex, Zena, and Orbis—each aligned to a specific market segment—and simplified performance tiers (Ultra, Premium, Pro, Nano, Pico), the new system helps partners and developers more easily map the right solutions to their needs. More than a branding update, this reflects Arm’s transformation from a core IP provider to a platform-first compute enabler, empowering faster innovation and alignment across the ecosystem. 

Inside Arm’s Innovations: Yellow Teaming methodology for building safer AI assistants

The Yellow Teaming methodology is reshaping how AI assistants are built—by proactively testing their safety, accuracy, and reliability before deployment. Zach Lasiuk, Principal Solutions Designer at Arm, explains how this approach strengthens LLM-based systems against real-world challenges—ensuring they meet the growing demands for trustworthy, secure AI.

Arm supports this evolution by delivering the scalable, power-efficient compute needed to run AI workloads seamlessly across everything from data centers to edge devices. 

Arm Neoverse powers scalable AV development with OpenAD kit

The OpenAD Kit is now deployable on Arm® Neoverse™ platforms, creating a powerful foundation for scalable, real-time autonomous driving software development. Odin Shen, Principal Solutions Architect at Arm, explains how the OpenAD Kit leverages Arm Neoverse to support seamless deployment across cloud and edge environments.

This enables performance-optimized simulation and validation workflows, accelerating the development and deployment of autonomous vehicle (AV) software. Arm provides the scalable compute infrastructure—from cloud to edge—that makes OpenAD Kit a key enabler of the future software-defined vehicle. 

How Arm brings realistic cloth simulation to mobile devices

Arm is pushing the boundaries of mobile graphics with a neural-graphics-powered cloth simulation that brings realistic, physics-inspired clothing movement to even low-power smartphones. Mina Dimova, Principal Engineer at Arm, explains how advanced cloth simulation—once limited to high-end platforms—is now possible on mobile using Arm Mali GPUs and efficient rendering techniques.

At the core is a mobile optimized graph attention network that replicates lifelike fabric behavior—folds, drapes, and collisions—without the heavy compute demands of traditional physics engines. This breakthrough allows developers to deliver console-quality visuals on mobile and XR devices, while staying within strict power and thermal budgets, showcasing Arm’s leadership in scalable, next-gen graphics solutions. 

Built on Arm: SOAFEE

SOAFEE (Scalable Open Architecture for Embedded Edge) is an Arm-led initiative redefining how automotive software is developed and deployed. Backed by over 150 partners, it brings together open standards, real-time requirements, and cloud-native tools to streamline the path to scalable, safety-critical automotive software.

In this video, Suraj Gajendra, VP Products and Solutions for Automotive at Arm and Governing Body Chairperson of SOAFEE, explains how SOAFEE is enabling a new era of software-defined vehicles through standardization, virtualization, and cloud-native development. Suraj also explains how virtual prototyping, early access to compute subsystems (CSS), like Arm Zena CSS, and open collaboration are helping OEMs and developers tackle rising complexity, reduce time-to-market, and embrace AI-driven innovation – before silicon is even available.

Godot Engine Team accelerates Vulkan with Arm’s developer tools

Godot 4 introduces a powerful new Vulkan renderer that shines on desktop platforms—but early mobile performance revealed hardware-specific challenges. Clay John, Rendering Maintainer and Board Member for Godot Engine, shares how the Godot Engine team tackled these issues by optimizing Vulkan for mobile GPUs—especially those built on Arm architecture.

Using Arm’s suite of developer tools—including Performance Studio, Streamline, and the Mali Offline Compiler—they identified and resolved critical performance bottlenecks in 3D scene rendering. The result: smoother, more efficient mobile gameplay, and a clear example of how Arm’s ecosystem empowers open source developers to fully harness their hardware’s capabilities. 

New performance libraries and toolchain boost Linux workloads on Arm 

The latest releases—Arm Performance Libraries 25.04 and Arm Compiler for Linux 20.0—bring advanced math optimizations and toolchain enhancements tailored for Arm-based systems.

Chris Goodyer, Director of Technology Management at Arm, highlights how these updates boost performance, improve code portability, and streamline developer workflows for both high-performance computing and cloud-native applications. With these releases, developers can expect faster, more reliable code execution on Arm hardware, powered by Arm’s foundational software tools that support scalable, energy-efficient innovation. 

How AI understands your input: tokenization explained

Curious how AI makes sense of language, images, or audio? This short video demystifies tokenization—the crucial first step that breaks down your input into data AI models can process. From words to pixels, discover how models like ChatGPT convert raw information into meaningful insight. It’s a quick, clear look at the mechanics behind modern AI.