The AI software portfolio, which has been designed for OEM and developers, works across the company-powered devices, spanning the range of connected intelligent edge products that include mobile, compute, automotive, extended reality, Internet of Things, and cloud platforms
Qualcomm has unveiled the new Qualcomm AI Stack portfolio. (Credit: Qualcomm Technologies, Inc.)
Qualcomm Technologies has launched a new unified artificial intelligence (AI) solution called Qualcomm AI Stack for original equipment manufacturers (OEMs) and developers to support a broad range of intelligent devices with AI software access and compatibility.
The AI software portfolio works across the company-powered devices, spanning the range of connected intelligent edge products that include mobile, compute, automotive, extended reality (XR), Internet of Things (IoT), and cloud platforms.
According to the semiconductor major, Qualcomm AI Stack supports various AI frameworks.
The solution is also compatible with well-known runtimes such as TensorFlow, ONNX, and PyTorch as well as developer libraries and services, tools, system software, and compilers.
Qualcomm Technologies product management vice president Ziad Asghar said: “This comprehensive Qualcomm AI Stack portfolio is a truly revolutionary step in improving AI development for OEMs and developers and achieving high performance across our product portfolio.
“We’re delighted to extend our one technology roadmap with this new portfolio and expand our leadership in the Connected Intelligent Edge.”
Qualcomm AI Stack offers direct access to the Qualcomm AI Engine as well as dedicated AI cores on Qualcomm Cloud AI 100 using Qualcomm AI Engine direct.
Qualcomm AI Engine direct, as part of the new solution, will now scale across all the AI accelerators inside a comprehensive variety of the company’s products.
It is an AI library which assigns and deploys currently available models directly to the AI accelerators on Qualcomm Technologies’ platforms.
Qualcomm AI Engine direct will now enable OEMs and developers to develop a feature and then transfer the same model across various products and tiers.
The Qualcomm AI Stack portfolio will also have a suite of tools such as Qualcomm AI Model Efficiency Toolkit (AIMET), AI development graphical user interface (GUI), model analysers to promote improved quantisation and optimisation, and neural architecture search (NAS).
Microsoft AI frameworks architecture engineer Yuan Yu said: “Qualcomm Technologies continues to deliver advanced AI solutions that make it easier for developers to leverage today’s most powerful AI and ML capabilities.
“We’re enthusiastic to bring Qualcomm AI Software technologies to developers in our ecosystems, including the Qualcomm Neural Processing SDK for Windows, to offer edge-to-cloud AI capabilities by providing unified, comprehensive AI software that scales from clients to servers.”