xMEMS Labs has developed a solid-state micro fan so small it can slip inside the temple arm of ordinary-looking smart glasses, and the company claims it can lower component temperatures by as much as 10°C. The XMC-1200, announced alongside a dedicated driver chip, is a piezoelectric MEMS device that consumes just 70mW while actively moving air past hotspots. If the technology delivers on its promises in actual products, it could solve one of wearable computing's most stubborn problems: how to sustain AI, video capture, and bright displays without making glasses hot enough to become uncomfortable after a few minutes.
What the XMC-1200 Actually Does
The chip measures 46 square millimeters and is about 1mm thin, making it uniquely suited for eyeglass frames where even the smallest rotary fan would be too thick, noisy, and power-hungry. Instead of a motor and spinning blades, it relies on piezoelectric membranes that flex at ultrasonic frequencies to push air through microscopic valves. The result is forced-air cooling without moving parts that can wear out—at least in the conventional sense.
Paired with xMEMS' Astra2 drive ASIC, the XMC-1200 moves up to 10 cubic centimeters of air per second directly across a targeted hotspot. The company’s internal measurements show up to a 10°C drop under a 1W thermal load, which is roughly the sustained power draw of a modern smart glass processor, camera pipeline, and always-on AI co-processor running simultaneously. That’s enough headroom to keep a system from throttling during extended use.
Importantly, the XMC-1200 is being designed for real-world wear: it carries an IP68 rating, so it can survive sweat, rain, and dust, provided the overall device design does not compromise the venting or seals. The fan’s solid-state nature also means no bearing noise, and because the actuation frequency is above human hearing, the cooling itself should be silent—though airflow through vents can still create subtle tones that engineers will need to manage.
XMC-1200 engineering samples are already with selected manufacturers under NDA, and the company aims for production readiness in the fourth quarter of 2027. That means commercial products using the chip are unlikely to appear before 2028 at the earliest.
Where Active Cooling Matters for Windows and Beyond
For Windows-focused readers, the most direct impact is on enterprise and industrial smart glasses that connect to Windows PCs, virtual desktops, and line-of-business applications. Devices like Microsoft’s HoloLens demonstrated that mixed reality can transform remote assistance, guided assembly, and field service, but they also highlighted the penalty of heat buildup during long shifts. A factory worker wearing camera-equipped glasses while streaming video and receiving annotated instructions may run into thermal throttling after a half hour, causing frame drops or app slowdowns. By keeping key components cooler, the XMC-1200 could make those workflows viable across a whole shift.
Consumer AI glasses, should they eventually run Windows-connected companion apps or integrate with Microsoft’s cloud AI services, would benefit similarly. Sustained on-device translation, object recognition, and visual search demand that neural processors and image pipelines run at full tilt for minutes, not seconds. Without active cooling, manufacturers are forced to offload those tasks to a phone or the cloud—adding latency and eating into the very responsiveness that makes AI eyewear compelling.
Even outside the Windows ecosystem, the same physics apply. Camera-equipped glasses that people actually want to wear for hours need to stay within a narrow skin-contact temperature envelope. The forehead and temples are far less tolerant of warmth than a phone in a pocket. A 10°C reduction at the source can mean the difference between a comfortable all-day device and one that gets put back in its case after a brief demo.
The Road to Cooler Wearables: How We Got Here
Google Glass first showed the world that a head-worn computer could be practical, but its thermal limits were as much a barrier as its price and social awkwardness. Since then, semiconductor efficiencies have improved drastically, but manufacturers have used those gains to add more demanding features rather than simply lowering power consumption. Every new generation of hardware now packs in larger cameras, brighter displays, louder speakers, and dedicated AI silicon, all of which generate heat in a frame that has almost no surface area for passive dissipation.
Passive strategies like graphite spreaders, thermal vias, and copper traces inside the temple arms can move heat around, but they cannot eliminate it. Eventually, warmth reaches the skin. Some designs try to push compute into an external puck or tether to a phone, but that defeats the stand-alone convenience that makes smart glasses appealing.
xMEMS is not the first to attempt active cooling in wearables. Traditional micro blowers and diaphragm-based pumps have been proposed before, but they either consumed too much power, created audible noise, or diluted the form factor with bulky enclosures. The company’s earlier XMC-2400 chip, a larger sibling already in production for smartphones and tablets, proved that piezoelectric cooling could work at scale. The XMC-1200 shrinks the concept to fit in eyeglass temples, a form factor where every tenth of a millimeter matters.
What This Means for You: Buy, Wait, or Watch?
If you’re an IT decision-maker evaluating smart glasses for your organization, the arrival of practical active cooling changes your timeline. Right now, most enterprise wearables will throttle under sustained mixed-reality or AI workloads. When you test devices today, ask vendors how long their glasses can record video or run an AI model before frame rates drop or the device becomes uncomfortably warm. That will become a key buying criterion once cooled products ship. Plan your deployment roadmaps with the expectation that by late 2028, active cooling might be a standard differentiator.
For everyday users curious about snapping up a pair of AI glasses, patience is in order. The first consumer products using the XMC-1200 are at least two years away. In the meantime, existing models that rely entirely on passive cooling should be judged by their sustained performance, not just their peak specs. If you notice a device getting hot after 10 minutes of video recording or live translation, that’s a sign the thermal design is at its limit.
Developers building apps for Windows or Android-based smart glasses should keep an eye on the integration kits and APIs that will appear once OEMs adopt the XMC-1200. The ability to tap into sensor-driven cooling policies—ramping the fan during heavy AI inference, then falling silent—could become a platform-level feature that separates high-end experiences from budget implementations.
Finally, if you own a pair of existing glasses, there’s no retrofit option. The XMC-1200 must be designed into a product from the start, with carefully engineered airflow paths and vent placement. This isn’t a drop-in heat sink; it’s a system-level component.
What to Watch Next
The most critical milestone isn’t the chip itself but the first reference designs from ODMs that show how to integrate it into a functional pair of glasses. Look for those prototypes at trade shows in late 2026 or early 2027. Independent thermal measurements conducted on actual prototypes—not lab jigs—will reveal whether that 10°C drop holds up in a real device, and whether skin-contact temperatures actually improve.
Also watch the acoustic performance. Even if the fan itself is inaudible, the sound of air moving through narrow temple vents, especially near the ear, could become a nuisance. Early feedback from partners will be telling.
Finally, the XMC-2400’s track record in shipping phones will foreshadow the XMC-1200’s reliability. If the larger chip proves durable and efficient in millions of handsets, confidence in the eyeglass variant will soar. The smallest component in the system may well become one of the most important for turning a hot, throttled wearable into a cool, persistent AI companion.