Mosaic SoC Raises $3.8m to Power Smaller Edge Devices

Mosaic SoC has raised $3.8 million in pre-seed funding to develop perception chips for small, battery-powered devices that need real-time spatial awareness without the energy demands of larger compute systems.

The Zurich-based company is targeting wearables, AR glasses, smartphone cameras and other edge devices where continuous environmental sensing is becoming more important, but where battery life, heat and physical size remain hard limits.

The funding round was led by Founderful, with participation from Kick Foundation. It comes as device manufacturers look for ways to add more intelligence to compact consumer hardware without depending on application-class processors, GPUs or power-heavy sensor processing stacks.

Why spatial computing needs more efficient hardware

Spatial computing relies on devices that can interpret the world around them, not just capture images. Smart glasses, mobile cameras and always-on AI features need to process visual and positional data in real time, building awareness of location, movement and nearby objects.

That requirement creates a hardware problem. Many existing systems use processors and graphics hardware that are too power-intensive for lightweight devices designed to run throughout the day. For products such as AR glasses, the challenge is not only performance, but whether the technology can fit into a form factor people would actually wear.

Mosaic SoC is developing integrated circuits intended to provide a lower-power layer of spatial intelligence. Instead of adding more compute-heavy components, the company’s approach is based on a dedicated perception chip that can process sensor data locally and reduce reliance on larger processing chains.

In practical terms, this could support smart glasses that map a room, recognise objects or help users locate items without requiring bulky hardware or frequent charging. For smartphones, Mosaic SoC sees its chip operating as a co-processor for the front camera, enabling features such as event-triggered recording, object-based activation or lower-power tracking.

From chip supplier to edge AI platform

Mosaic SoC was founded by ETH Zurich PhDs Moritz Scherer and Alfio Di Mauro, who identified a gap between the growing demand for edge intelligence and the hardware available to support it.

The company’s initial business model centres on selling integrated circuits, but its wider strategy includes AI deployment tools, compilers and developer workflows for firmware teams. This could make integration easier for original design manufacturers that want perception capabilities without building the full software and hardware stack from scratch.

Its early revenue has come from non-recurring engineering contracts with ODM partners. Over time, Mosaic SoC expects more of its revenue to come from scalable chip sales as customer products move toward market.

The company will still need to prove that its architecture can be manufactured reliably, priced competitively and integrated within the product cycles of consumer electronics brands. That is a high bar in a market where performance claims are only useful if they translate into shipping hardware.

For the wider electronics sector, Mosaic SoC’s funding reflects a clear trend: more AI and sensing workloads are moving closer to the edge. The next phase of smart devices may depend less on whether they can collect data, and more on whether they can interpret it continuously, efficiently and quietly enough to become part of everyday use.