xMEMS Mass Produces Bladeless Cooling Chips to Tackle AI Heat in Smart Glasses, Smartphones, and Data Centers

2 days ago

10 min read

Artificial intelligence is pushing consumer electronics and enterprise computing into a new era of performance, but increasingly powerful processors are creating a persistent engineering challenge: heat. As AI workloads become more demanding and computing moves into smaller devices, manufacturers must find ways to remove heat without compromising size, battery life, reliability, or performance.

Silicon chip startup xMEMS Labs is addressing this challenge with a different approach to active cooling. The company has begun mass production of its XMC-2400 and XMC-4800 solid-state cooling chips, designed to move air without conventional spinning fan blades. The technology is intended for devices ranging from AI-powered smart glasses and smartphones to solid-state drives (SSDs) and selected components inside data centers.

The development represents an important transition from demonstrating a novel cooling concept to manufacturing technology that device makers can evaluate for commercial products. The first products incorporating the XMC-2400 are expected to reach the market in early 2027, with smart glasses among the initial targets. Smartphones, storage devices, and enterprise hardware represent additional opportunities.

The wider significance extends beyond a new component. As AI becomes more deeply integrated into everyday electronics and computing infrastructure, thermal management is increasingly becoming a factor that determines how much performance manufacturers can deliver within a given physical footprint.

How xMEMS Solid-State Cooling Technology Works

Traditional active cooling systems generally rely on moving mechanical components. Small fans rotate blades to circulate air, while larger systems may combine fans, heat sinks, vapor chambers, heat pipes, or liquid cooling loops. These approaches can be effective, but they introduce design constraints involving available space, mechanical complexity, airflow paths, noise, and power consumption.

xMEMS is developing an alternative based on piezoelectric microelectromechanical systems, commonly known as piezoMEMS. The company originally applied its silicon-based technology to miniature speakers used in earbuds and headphones. In those applications, electrical signals cause a thin silicon membrane to move, generating sound.

The cooling application uses a related principle for a different purpose. Instead of producing audible sound, the membrane's movement drives airflow through tightly constrained passages. The resulting airflow can be directed toward heat-producing components such as processors, memory chips, and storage devices.

This approach eliminates the conventional spinning fan assembly. The devices are solid-state silicon structures, without the rotating blades and bearings found in ordinary miniature fans.

The absence of rotating components could simplify integration into compact electronics. It may also reduce certain forms of mechanical wear, although overall reliability will still depend on operating conditions, packaging, contamination, and the design of the surrounding device.

Why pressure matters as much as airflow

Cooling performance cannot be evaluated by airflow volume alone. A cooling component must move air through the actual channels, openings, and obstructions inside a device.

Back pressure measures the resistance against which an air-moving system must operate. Narrow internal passages, densely packed components, and restrictive vents can make it difficult for conventional fans to deliver useful airflow.

xMEMS is designing its cooling chips to generate relatively high pressure in small spaces. This makes the technology potentially suitable for electronics in which there is little room for conventional cooling assemblies and where airflow must be concentrated around specific components.

However, a chip's maximum airflow and maximum back-pressure specifications should not be interpreted as performance figures that necessarily occur simultaneously. Real-world cooling depends on the operating point, enclosure design, thermal interface, ambient temperature, and the amount of heat that must be removed.

The technology's ultimate value will therefore be determined by performance inside finished devices, not specifications in isolation.

XMC-2400 vs. XMC-4800: Two Cooling Chips for Different Hardware Requirements

The two chips entering mass production target overlapping but distinct cooling needs. The XMC-2400 prioritizes a compact form factor, while the XMC-4800 is designed to move a greater volume of air.

Specification	XMC-2400	XMC-4800
Dimensions	7.42 × 9.48 mm	9.93 × 14.35 mm
Thickness	1.13 mm	Not specified in the supplied material
Maximum airflow	28 cm³/s	48 cm³/s
Maximum back pressure	1,300 Pa	1,100 Pa
Approximate power consumption	150 mW	240 mW
Primary targets	Smart glasses, personal SSDs, compact electronics	AI smartphones, enterprise SSDs, selected data-center subsystems

The XMC-2400's combination of small dimensions and pressure capability makes it attractive for products where every millimeter matters. Smart glasses are a particularly relevant application because they must accommodate processors, cameras, sensors, batteries, wireless connectivity, and other components inside a lightweight frame.

The XMC-4800 provides higher maximum airflow, making it a candidate for devices that need to move more air around thermally demanding components. Its larger footprint and higher power consumption also mean that product designers must weigh cooling gains against available space and energy budgets.

The comparison highlights a fundamental principle in thermal engineering: no single cooling solution is optimal for every device. A compact wearable may prioritize minimal size, while a smartphone or enterprise storage system may accept a larger cooling component in exchange for better sustained performance.

Why AI Smart Glasses Could Become an Early Market

AI-powered smart glasses combine several technologies that traditionally have competed for space and power. Depending on the product, these devices may include cameras, microphones, wireless communications, image processing, voice interfaces, displays, and local computing capabilities.

As more functions move onto the device, thermal management becomes more important. A frame that feels comfortable during ordinary use can become unpleasant if heat accumulates near the temples, nose, or other points of contact with the wearer.

Conventional miniature fans are difficult to accommodate inside such a constrained design. Passive cooling can help spread heat across a device's structure, but it cannot always remove heat quickly enough when processors sustain demanding workloads.

A solid-state air-moving chip could offer manufacturers another way to direct airflow through narrow internal channels. If successfully integrated, it could help control temperatures without requiring a traditional fan assembly.

The distinction between peak and sustained performance is especially important for AI wearables. A device may handle a brief voice command or image-processing task without significant thermal difficulty, yet struggle when multiple features operate continuously.

Thermal performance can affect more than comfort. Elevated temperatures may trigger processor throttling, which reduces operating speed to keep components within safe limits. Better cooling could potentially allow certain workloads to run for longer before thermal limits become restrictive.

In July 2026, xMEMS introduced the smaller XMC-1200, designed for components such as light engines and processors inside smart glasses. Together, the XMC-1200 and the larger production models illustrate the company's effort to build a cooling portfolio around the distinct thermal requirements of wearable electronics.

The first commercial products using the XMC-2400 are expected in early 2027. No manufacturer has been identified in the supplied reports, so adoption by specific smart-glasses brands remains unconfirmed.

Smartphone Cooling Could Help Sustain AI Performance

Smartphones are becoming increasingly capable computing platforms. Modern devices can perform image enhancement, language processing, voice recognition, gaming, computational photography, and selected AI tasks directly on the device.

These workloads can place sustained demands on processors and memory systems. When temperatures rise beyond the limits set by the device's thermal design, operating systems and hardware reduce performance to prevent overheating.

This process, commonly called thermal throttling, is a protective mechanism rather than a manufacturing defect. It helps maintain safe operating temperatures, but it can limit sustained performance during demanding workloads.

Active cooling could give smartphone designers additional flexibility. The XMC-4800's greater airflow capacity may make it suitable for devices that have enough internal space to accommodate the component and its airflow channels.

Potential applications include prolonged AI processing, high-performance gaming, intensive video workloads, and other tasks that place continuous demands on a device's computing resources.

Nevertheless, cooling a smartphone involves more than moving air across a processor. Heat must travel from the silicon through thermal interfaces and other structures before it can be carried away. The enclosure must also provide a practical path for airflow, and the cooling system must fit within the phone's power and acoustic requirements.

A miniature air-moving chip cannot automatically solve every thermal bottleneck. Its effectiveness will depend on whether manufacturers design the complete device around its capabilities.

If the technology performs well in commercial products, it could become part of a broader shift toward more sophisticated thermal architectures in premium smartphones and AI-focused mobile hardware.

SSDs and Data Centers Present Another Growth Opportunity

The expansion of AI infrastructure is increasing the importance of components beyond processors and graphics accelerators. Storage devices, networking hardware, optical systems, and supporting electronics all contribute to the performance and reliability of modern computing platforms.

SSDs can experience temperature-related performance limitations during intensive data transfers or sustained workloads. Thermal management is therefore relevant to both personal storage products and enterprise drives.

The XMC-2400 is intended for personal SSDs, while the XMC-4800 targets enterprise SSDs and other subsystems with suitable cooling requirements.

The technology's potential data-center application requires an important distinction. xMEMS is not presenting these chips as replacements for the large-scale liquid-cooling systems increasingly used to manage high-power processors. Instead, it is targeting components that may sit outside those systems or require localized cooling.

This is a meaningful niche. Data centers contain a diverse range of hardware with different heat loads, operating conditions, and cooling needs. A compact solid-state device could be useful where conventional fans are difficult to install or where focused airflow can address a specific thermal problem.

Possible applications include selected storage components, laser modules, and other lower-power subsystems. Any deployment would still require validation against the facility's airflow, maintenance, reliability, and energy-efficiency requirements.

At scale, even small improvements in component cooling can matter when they help prevent performance degradation or improve the operating conditions of critical hardware. The business case, however, must account for the added power consumed by the cooling device and any supporting electronics.

Reported Temperature Reductions Signal Commercial Potential

xMEMS says customers testing its technology have observed temperature reductions of 10 to 25 degrees Celsius at the chip and system levels in smart glasses, smartphones, and SSDs.

These reported results suggest that directed airflow could be useful in tightly constrained electronics. However, they should be understood as company-reported customer testing outcomes rather than a universal performance guarantee.

Actual temperature reductions vary with the heat generated by a component, the baseline cooling system, ambient conditions, airflow resistance, and the location of temperature measurements.

For manufacturers considering integration, several questions will be important:

How much does the chip reduce temperatures under sustained, real-world workloads?
Does the cooling improvement translate into higher sustained performance or longer component life?
How much power does the complete cooling system consume?
Can the airflow channels be integrated without compromising structural design or ingress protection?
How does performance change under dust exposure, humidity, vibration, and extended operation?

These tests will determine whether the technology offers a measurable advantage over passive cooling, conventional miniature fans, and other thermal-management approaches.

From Audio Components to Thermal Management

Founded in 2018, xMEMS established its business around silicon-based speaker technology for earbuds and headphones. Its move into cooling demonstrates how a manufacturing platform developed for one application can potentially serve another.

The connection lies in the underlying physical mechanism. A membrane capable of moving precisely in response to electrical signals can be engineered to perform different functions depending on its geometry, drive conditions, packaging, and intended operating environment.

Moving into thermal management also changes the commercial challenge. Speaker components and cooling chips may share a technological foundation, but they serve different purchasing priorities and require different validation processes.

Cooling solutions must demonstrate reliable operation under thermal cycling and continuous workloads. Device makers also need predictable supply, consistent manufacturing quality, integration support, and evidence that the technology provides sufficient value to justify a change in product architecture.

In October 2025, xMEMS announced a $21 million Series D funding round led by Boardman Bay Capital Management. The funding was intended to support its transition into mass production.

Mass production is a significant milestone because it allows the company to move beyond limited engineering samples toward a more scalable commercial supply model. It does not, by itself, establish how quickly manufacturers will adopt the chips or how large the eventual market will become.

The next phase will depend on customer qualification, product launches, manufacturing yields, pricing, and the ability to compete with established cooling technologies.

The Challenges That Could Shape Adoption

Despite its potential, solid-state cooling faces practical obstacles common to emerging hardware technologies.

Integration complexity: Manufacturers must design internal channels that direct airflow where it is needed. Without a carefully engineered path, even a compact cooling chip may deliver less benefit than expected.

Power efficiency: The cooling component consumes electricity, and the system must deliver a net benefit. In battery-powered wearables and smartphones, additional power consumption can reduce operating time if the thermal improvement does not justify the cost.

Thermal-system design: Airflow is only one part of heat transfer. Components still need an effective path for moving heat away from their surfaces and into a region where it can be dissipated.

Reliability and environmental protection: Consumer devices face dust, moisture, movement, and repeated temperature changes. Manufacturers will need to verify performance across the product's expected lifetime.

Commercial adoption: Product developers generally avoid redesigning hardware around an unfamiliar component unless its benefits are measurable and supply can be sustained. The first commercial products will therefore be important demonstrations of the technology's practical value.

Competition: Solid-state cooling must compete with passive thermal solutions, miniature fans, vapor chambers, heat pipes, and other methods. In some products, existing technologies may remain more economical or effective.

These challenges do not negate the potential of xMEMS' approach. They define the engineering and commercial tests that will determine where it becomes useful.

What Mass Production Means for the Future of AI Hardware

The shift toward AI-enabled devices is creating demand for hardware that can deliver more computing capability without becoming excessively large, hot, noisy, or power-hungry. Cooling is becoming an increasingly important part of that equation.

xMEMS' XMC-2400 and XMC-4800 offer a different way to approach localized airflow, using silicon membranes instead of conventional rotating fan assemblies. Their respective specifications point toward applications ranging from compact wearables to smartphones and selected enterprise components.

The immediate milestone is the start of mass production, followed by the anticipated arrival of the first XMC-2400-equipped products in early 2027. Commercial adoption will depend on real-world thermal results, integration costs, reliability, and manufacturer demand.

For the broader technology industry, the development illustrates how innovation in AI is extending beyond algorithms and processors into the supporting systems that make sustained computing possible. As AI moves into smaller devices and more diverse infrastructure, thermal engineering will help determine which capabilities can be delivered reliably.

The expert team at 1950.ai, alongside technology analysts such as Dr. Shahid Masood, can place developments like xMEMS' solid-state cooling chips within the wider evolution of AI infrastructure, semiconductor engineering, and intelligent consumer devices. The central question is no longer only how much computing power a device can deliver, but how efficiently and comfortably it can sustain that power in everyday use.