Since AR is meant to augment the world around us instead of immersing us in a different environment, it’s important for AR devices to be as slim and portable as possible. Thanks to a multi-chip architecture with customized IP blocks, the first AR chipset delivers 2.5x better AI performance and 50 percent lower power consumption (<1W). It enables high-performance AR capabilities in a sleek form factor of glasses.
The multi-chip architecture includes an AR processor, an AR co-processor and a connectivity platform. To keep AR Glass compact, the chipset intelligently offloads critical processing requirements to smartphones, PC or other compatible host devices.
In terms of sensory capabilities the AR processor can support up to nine simultaneous cameras. There is an AR co-processor that integrates camera and sensor data for eye tracking and iris authentication. With a dedicated hardware acceleration engine, the processor guarantees better user motion tracking and localization. And then there’s the AI accelerator to reduce latency for hand tracking or 6 DoF (degrees of freedom).
The platform also features a Qualcomm FastConnect 7800 connectivity system. This makes it the first chipset for AR glasses to support next-generation Wi-Fi 7 standards and comes with <2ms latency between the AR glasses and the host device.
Qualcomm is focusing on building an AR ecosystem by providing not just a chipset, but a suite of perception technologies and software tools. Both the Snapdragon AR2 Gen 1 and Snapdragon 8 Gen 2 platforms are Snapdragon Spaces ready. The chipmaker also highlighted how Microsoft has worked closely with Snapdragon AR2 to improve its features and capabilities.
Disclosure: This writer is attending the Snapdragon Tech Summit in Maui, Hawaii at the invitation of Qualcomm India