Qualcomm AI Hub Expands to On-Device AI Apps for Snapdragon-Powered PCs

(Auszug aus der Pressemitteilung)

  • Qualcomm AI Hub expands to support Snapdragon X Series Platforms, empowering developers to easily take advantage of the best-in-class CPU and the world’s fastest NPU for laptops, and create responsive and power-efficient on-device generative AI applications for next-gen Windows PCs.
  • Developers can now optimize their own models using the Qualcomm AI Hub—adding flexibility and customization while accelerating development cycles—in addition to utilizing its growing library of pre-optimized AI models.

At Microsoft Build 2024, Qualcomm Technologies, Inc. announced the expansion of its Qualcomm® AI Hub to support Snapdragon® X Series Platforms—reducing time-to-market for developers and unlocking the benefits of on-device generative AI in next-gen Windows PCs. Additionally, developers can now bring their own AI models and seamlessly optimize them to run directly within devices. The ability to run AI models on devices offers significant advantages, including superior responsiveness, enhanced privacy, improved reliability, and more personalized experiences for users.

“With support for Snapdragon X Series Platforms in the Qualcomm AI Hub, we are providing developers with the flexibility and power to create innovative AI applications for next generation Windows PCs,” said Durga Malladi, senior vice president and general manager of technology planning and edge solutions, Qualcomm Technologies, Inc. “We are committed to providing developers with the tools and resources they need to unleash the full potential of on-device generative AI.”

The Qualcomm AI Hub is the developer’s gateway for resources, tools, and services to enable superior on-device AI performance and features a growing library of more than 100 pre-optimized AI models, which are also available on Hugging Face and GitHub. Additionally, Qualcomm AI Hub now allows developers to “Bring Your Own Model” (BYOM), with the capability to upload, optimize and compile their models specifically for Snapdragon and Qualcomm® platforms – including support for all key frameworks such as PyTorch, TensorFlow, and ONNX. Developers can easily test and validate the models on cloud-hosted devices, in less than five minutes and using a few lines of code. By simplifying the model optimization and verification process, developers can save valuable time and effort, accelerating their development and deployment cycles facilitating faster iterations.

Leading global PC OEMs, including Acer, ASUS, Dell, HP, Lenovo, Microsoft Surface, and Samsung are scheduled to launch PCs powered by Snapdragon X Elite and/or X Plus Compute Platforms, allowing developers to use the Qualcomm AI Hub to bring superior experiences to this new generation of devices.

Snapdragon X Series Platforms support a broad range of AI development tools. Thanks to deep integrations with ONNX Runtime and Hugging Face Optimum, as well as support for Llama.cpp, developers can now run thousands of models on Snapdragon X Series Platforms and build rich AI-powered Windows applications. Developers can also learn about on-device AI and how to implement it for their apps in this online course offered by DeepLearning.AI in collaboration with Qualcomm Technologies. For updates, subscribe to our monthly developer newsletter.