Can smartwatches and glasses get smarter? Inside Qualcomm and MediaTek’s new wearable AI game

With all the talk about superchips driving AI inference in the data center and the Spark RTX architecture taking Windows AI capabilities to unprecedented heights, it’s easy to overlook the room for improvement in some of the most challenging device classes in power management: IoT and wearables.

For products such as environmental sensors that may need to relay information from remote areas, battery life measured in years may be essential. On the other hand, small wearable devices such as rings, watches, and pendants require smaller batteries while their biomarker monitoring functions

However, two of the largest SoC vendors have turned their attention to one of these markets with major upgrades to their product lines.

Genio prepares for physical AI

Mediatek has long targeted the IoT market, which includes a mix of applications such as digital signage, smart city sensors, retail and industrial robotics. The latter is increasingly moving toward physical AI, where the intelligence of machines can match their ability to lift and move objects.

The company seeks to accelerate the market with its new Genio Pro 5100 IoT chip, built on an industry-leading 3nm process, combining CPU performance that rivals Intel’s Core chips, ARM’s Immortalis GPU architecture, and a powerful NPU.

With many IoT applications focused on image analysis, the chip can support up to 16 Full HD camera inputs or/or two 4K inputs (both at 30 frames per second) as well as platform-optimized Vision Langauge models.

Robotics is one of the fastest growing applications for this type of chip. MediaTek supports both Linux and ROS (Robotics Operating System) for applications in industrial automation, transportation, logistics and hospitality and intends to support the line for seven to 12 years.

Although the Genio Pro doesn’t include wireless connectivity, modules are available for Wi-Fi and Bluetooth as well as cellular, including the low-power RedCap 5G.

A new generation of wearables

Qualcomm has been active in the wearables market for years (and even launched its own smartwatch called Toq, although that was more aimed at showcasing its since-discontinued Mirasol display technology).

Variants of its W5 chip have powered the latest generations of Pixel watches. The company’s new Snapdragon Elite represents a significant increase in performance over the previous W5, which, like MediaTek’s current-generation Genio IoT chip, will remain in the more cost-sensitive device market.

The new SoC, which like the Genio Pro is based on a 3nm process node, features improvements in four main areas: micro-power connectivity, low-power islands, performance and user experience. While wearable devices should instantly apply AI processing to the inputs around us, Qualcomm has implemented a dual NPU architecture.

The main NPU can handle the same level of token processing as a two-year-old smartphone (around 18-20 tokens per second for a model with around 1 billion parameters). However, a new eNPU is designed to perform tasks that require rapid responses such as voice activation and active noise cancellation, while allowing days of use between charges.

The new chip also supports Qualcomm’s micropower Wi-Fi (which uses 80% less power than the previous generation), as well as the low-power 5G RedCap cellular standard and satellite connectivity.

This flexibility is intended to help drive the next wave of wearable applications such as health and fitness coaching, life blogging, on-device translation, and USB-based authentication using standards such as Aliro, a sister standard to Matter.

At Mobile World Congress, Samsung, Motorola, and Google announced support for the Snapdragon Wear Elite, with the former noting that the new chip will be used in at least one version of the Galaxy Watch 9.

However, we should see many more devices based on advancements from both companies in 2027 and beyond, bringing new levels of intelligence to the products that are constantly with us and many of which work invisibly on our behalf.

Ross Rubin is the founder and principal analyst of Reticle Research, a technology research and consulting company. Ross is an industry analyst specializing in product, service, and enabling infrastructure innovation in technology, media, and telecommunications markets for more than 30 years, writing columns for Engadget, ZDNet, and Fast Company, among other publications. You can contact him on LinkediIn and BlueSky.