Qualcomm Technologies recently announced the Qualcomm® AI Stack portfolio, accelerating the company’s leadership in Artificial Intelligence (AI) and the Connected Intelligent Edge.
Qualcomm AI Stack combines and enhances its best-in-class AI software offerings and is a comprehensive AI solution for OEMs and developers, supporting a wide range of intelligent devices with broad AI software access and compatibility. For the first time, a single AI software portfolio works for Qualcomm Technologies powered devices, spanning the broad range of Connected Intelligent Edge products, including mobile, automotive, XR, compute, IoT and cloud platforms.
Qualcomm AI Stack supports a variety of AI frameworks and popular runtimes, including TensorFlow, PyTorch, and ONNX, developer libraries and services, system software, tools, and compilers, so any AI feature developed for one device can be easily deployed to another.
This comprehensive portfolio provides direct access to the Qualcomm® AI Engine and dedicated AI cores on Qualcomm® Cloud AI 100 using Qualcomm® AI Engine directly. As part of the new offering, the Qualcomm AI Engine will now be scalable directly to any AI accelerator across a wide range of Qualcomm Technologies products. Qualcomm AI Engine direct is an AI library that directly delegates and deploys existing models to the AI accelerators on Qualcomm Technologies platforms and now gives OEMs and developers the ability to develop one feature and then deploy the same model across different products and levels. move.
The Qualcomm AI Stack portfolio will also feature a range of tools, including Qualcomm® AI Model Efficiency Toolkit (AIMET), AI development graphical user interface (GUI), model analyzers to promote enhanced quantization and optimization, and Neural Architecture Search (NAS). As announced at Snapdragon Tech Summit 2021, Qualcomm Technologies and Google Cloud have integrated Google Cloud Vertex AI NAS, enabling OEMs and the ecosystem to create, test and deploy unique and efficient experiences at the edge. The Qualcomm Neural Processing SDK continues to be an important product for OEMs and developers to run their neural networks on a wide variety of Qualcomm Technologies products and will continue to receive regular updates.test and support. This collective gives Qualcomm Technologies a foundation to create domain-specific SDKs for deployment in the Connected Intelligent Edge.