Qualcomm bets on AI wearables to replace the smartphone
Qualcomm is aggressively positioning its hardware as the foundation for the next generation of computing, betting that the smartphone will eventually be sidelined by a diverse ecosystem of AI-powered wearables. CEO Cristiano Amon revealed that the company is currently involved in over 40 distinct projects, ranging from smart jewelry and earbuds equipped with cameras to advanced pins and watches.
To support this shift away from traditional mobile handsets, the chipmaker unveiled two strategic offerings: the Snapdragon Reality Elite platform, aimed at mixed-reality headsets, and the Scalable Turnkey AI-Ready Toolkit (START), designed to accelerate the production of AI-integrated eyewear.
Advancing Mixed Reality Performance
The Snapdragon Reality Elite platform marks a significant technical leap over previous XR iterations. According to Qualcomm, the new architecture delivers performance gains of 60% in GPU, 30% in CPU, and 160% in NPU capabilities. These improvements are critical for running complex AI models directly on-device, with the company noting the platform can process a 3-billion-parameter language model at 45 tokens per second.
Beyond raw speed, the platform enhances the user experience through:
- Visual Fidelity: Support for 4.4K resolution per-eye at 90 fps, reducing motion sickness and eye strain.
- Enhanced Tracking: Improved head and hand tracking sensors for more precise interaction.
- Versatility: Capability to power both standalone video-see-through (VST) headsets and lightweight tethered optical-see-through (OST) glasses.
The technology is already gaining traction with hardware partners, including the XREAL Project Aura and an upcoming device from Play for Dream.
Lowering Barriers with the START Toolkit
With the introduction of the Scalable Turnkey AI-Ready Toolkit (START), Qualcomm aims to simplify the development process for manufacturers. The toolkit bundles AR-specific chips, software stacks, and companion applications, providing a white-label framework for faster market entry. The program currently offers three primary reference designs:
- Audio and camera-integrated frames similar to Meta’s Ray-Ban glasses.
- Monocular display units.
- Binocular display systems.
Eyewear manufacturers Inspecs and TitanFlex-owned O’Neill are among the first to adopt this white-label strategy, though Qualcomm intends to expand the toolkit to support various other form factors in the near future.
The Strategic Shift to AI Agents
Amon’s vision, as discussed with CNBC, centers on the necessity of real-world data to fuel AI agents. By moving beyond the smartphone, Qualcomm believes hardware startups will experiment with form factors that provide constant context—devices that can “see” the environment and facilitate seamless interaction with AI.
“The principle is something that you wear, something that is with you all the time, something that can see the world around you,” Amon stated. By providing the essential silicon layer for these new devices, Qualcomm is positioning itself to remain the backbone of personal computing, even as the dominance of established players like Apple and Samsung faces new competition from a wave of specialized hardware startups.