Arm has partnered with PyTorch to launch ExecuTorch 0.7, a runtime designed to run generative AI models efficiently on edge devices. This move aims to bring advanced AI capabilities, such as on-device chatbots and image generators, to smartphones, IoT devices, and other low-power hardware.
The new platform allows developers to deploy models like Llama and Stable Diffusion on Arm-based chips without relying on cloud servers. ExecuTorch 0.7 optimizes performance and memory usage, enabling real-time AI inference on devices with limited resources.
"This is a critical step in making generative AI accessible to billions of users," said an Arm spokesperson. "By running AI locally, we reduce latency and enhance privacy, as data doesn't leave the device."
The release includes pre-built examples and tools for model conversion, quantization, and deployment. Arm expects ExecuTorch to accelerate AI adoption across industries, from healthcare to automotive.