Real-time neural-camera denoising on Arm hardware using SME2 improves low-light photography and 4K video quality, enhancing camera pipelines on mobile devices.

Arm MCP Server extends cloud-migration tooling through AI-assisted workflows, enabling image inspection, code analysis, and architecture validation via natural language commands.

Arm Model Gym provides a developer toolkit to train and deploy neural graphics models, enabling neural supersampling and end-to-end workflows on Arm hardware.

Alif Semiconductor showcases embedded controllers and Ethos-U85-powered processors for intelligent edge computing and on-device AI in smart homes and wearables.

Vivo X300 series is highlighted as a flagship device built on Arm, with SME2 delivering AI acceleration across vision, speech, and text workloads.

Arm’s October 2025 innovations accelerate on-device AI, redefining edge intelligence for gaming, neural graphics, and embedded systems, with a focus on real-time processing, privacy, and richer user experiences.

The lineup highlights edge devices, automotive platforms, data centers, and creative tools, demonstrating advances in AI, efficiency, and scalability across platforms.

A breakthrough enables real-time on-device large language model inference on smartphones for mobile gaming, featuring Space Bartender with Whisper input and a 4-bit SmolLM2 model delivering improved performance with low power use.

Mobile LLMs run on-device for gaming, powered by Arm CPUs, KleidiAI, and Unreal Engine Neural Network Engine, demonstrated with Space Bartender and a 4-bit SmolLM2 model for real-time on-device inference.

Arm SME2 optimizations showcase boosted AI inference performance and scalability across modalities, leveraging Ethos-U NPU families.

ExecuTorch 1.0, developed with Meta, becomes generally available to enable low-latency, power-efficient AI on CPUs, GPUs, and NPUs across edge devices from microcontrollers to smartphones.

Google Cloud collaboration powers smarter, AI-defined vehicles by combining Arm compute with Google Cloud AI and data capabilities.