Google and Synaptics have released open-source AI chips designed for wearables and industrial devices. The chips are meant to boost what those devices can do while keeping battery drain in check. Both companies announced the collaboration this week, though they didn't put a date on when the first hardware using the designs will reach shelves.
What the chips are meant to do
The chips are built to run AI workloads directly on a device rather than sending data to the cloud. That's a big deal for wearables, where every milliwatt counts, and for industrial sensors that often sit in places where a stable network connection isn't a given. By handling inference on-device, the hardware can cut latency and keep sensitive data local. Google and Synaptics say the designs are open-source, which means other companies can take the reference architecture and build their own versions without starting from scratch.
Why open-source matters here
Open-sourcing chip designs isn't the norm. Most silicon vendors guard their blueprints closely. By putting the designs out in the open, Google and Synaptics are betting that a broader ecosystem will adopt the platform faster than a proprietary one would. For smaller hardware makers, that could lower the barrier to entry — they won't need to license a full stack or design a chip from the ground up. The catch is that open-source hardware still needs robust tooling and support, and neither company has said how much of that they'll provide.
The battery-life pitch
Battery life is the headline promise. Wearables like smartwatches and fitness trackers are constrained by tiny batteries, and adding AI features usually means either a bigger battery or more frequent charging. The new chips are designed to run AI models efficiently enough that device makers can add features without sacrificing days of battery life. Industrial devices have similar constraints — many are deployed in the field for years and can't be easily recharged or serviced. If the chips deliver on the efficiency claims, they could make always-on AI sensing practical in places where it wasn't before.
What's still missing
Google and Synaptics haven't released performance benchmarks, power figures, or a list of device makers that have signed on. They also haven't said whether the chips will be manufactured by Synaptics itself or through a foundry partner. Those details matter for anyone thinking about building on the platform. For now, the announcement is a design release, not a product launch. The first real test will come when third parties start shipping hardware based on the reference designs — and when independent tests can confirm the battery-life claims.
Neither company has set a public timeline for when that might happen.
