Qualcomm AI Hub: Optimized development options for on-device AI
27. November 2024 | Planegg
The new Qualcomm AI Hub is here to revolutionize the development of on-device AI!
With Stable Diffusion, Whisper, and LLaVA (available on platforms such as Hugging Face and GitHub) the new AI Hub provides a large selection of pre-optimized AI models, specifically designed for the development of on-device AI with Snapdragon and Qualcomm platforms.
The various models offer optimized features such as speech processing, text-to-image, object recognition and enable an improved developer and user experience thanks to faster, more secure and more cost-efficient on-device AI development for i.e. smartphones, PCs, IoT and automotive solutions.
The AI Engine's hardware acceleration across NPU, CPU and GPU, helps reduce inference times by a factor of four and allows for faster processing of individual steps.
Qualcomm AI Hub also supports Large Language Models (LLMs), such as the multimodal LLaVA model which enables the local processing of text and image input for a more optimized performance and increased data privacy. With Low Rank Adaptation technology (LoRA), AI hub enables stable diffusion on Android and thereby optimizes and simplifies the scaling and adaptation of tailor-made AI models for the using generative AI in image processing and language model adaptation.
Atlantik Elektronik supports you in integrating the Qualcomm AI Hub into your projects to develop locally optimized, secure and powerful AI solutions for a wide range of industries.
Learn more about the new Qualcomm AI hub and the options for your applications here!
