Overlord Labs Bets on a Grain-of-Rice Chip to Stretch Smart Glasses Battery Life

Wearables keep shrinking, but the battery hasn't. That mismatch is the problem Overlord Labs wants to solve, and the San Jose startup has raised $10 million to do it. Its pitch is a battery-intelligence platform built around a chip roughly the size of a grain of rice.

Here's the thing: most people never think about the battery gauge in their gadget until it misleads them. Overlord's argument is that battery management has been neglected for years, and that AI-powered smart glasses are exposing the gap.

A chip the size of a grain of rice

Overlord Labs is a fabless semiconductor company in San Jose, California. It builds what it calls battery-intelligence infrastructure for compute-intensive, battery-powered devices. Its Genesis platform combines precision sensing, programmable processors, software, and battery algorithms under the company's proprietary Silicon × Algorithm Architecture.

The hardware is the eye-catching part. The chip measures about 2.6 millimeters by 1.6 millimeters, with an indicative price of roughly $1. The firm manufactures it at South Korea-based DB HiTek on a mature 130-nanometer process, a choice that favors cost and yield over leading-edge density. For a part that sits quietly next to a battery, that tradeoff makes sense.

Working silicon is in hand, and customer sampling is underway with a leading U.S. AI glasses manufacturer. The company isn't naming that customer. Smart glasses are the first target, but the company sees the same logic applying to wearables, hearables, drones, and other compact devices.

The firmware angle deserves more attention than it usually gets. Support for firmware updates means battery models and control strategies can change as manufacturers add new workloads or move to new battery chemistries. A gauge that's fixed at the factory ages badly. One that updates can keep pace with how a device is actually used.

Why smart glasses make the battery problem worse

Smartphones carried battery management through the last decade, but the design stopped moving around 2012. Phones settled on big screens and big batteries, and much of the industry moved on. Battery chemistries kept improving, though, and the devices around the phone have much tighter constraints.

Cameras, audio, sensors, displays, and AI workloads all pull on the same cell, and each one changes power delivery, runtime, and the choices designers can make. Thermal conditions matter. So does battery placement. Put the cell in the wrong spot and the device gets heavier or less comfortable, which is exactly what smart glasses can't afford.

Richard Nicholson, Overlord's CEO, pointed to the Snap glasses as a clear example. They're still bulkier than ordinary eyewear and have to render complex augmented reality graphics on their optical lenses. "Everybody is struggling to get features and runtime into the devices," he said. "It's hard to see people wearing some of the devices all day long."

Carlos Rodriguez, the company's COO, makes the point more bluntly. The battery is the largest physical element in Meta's market-leading AI glasses, and it dictates how big the frame has to be. He wears a pair himself and says they deliver about four hours of use. To get through a day, he carries a second pair. AI features make that worse because they're needed all the time, not just when someone opens an app.

Gauges, margins, and wasted capacity

Here's where the engineering gets practical. Without a precise reading of remaining charge, device makers hedge. Overlord says engineers often leave a 30% power margin just to make sure a device doesn't die early. That's a lot of capacity sitting unused, and in a device the size of a pair of glasses, it's capacity the designer could have spent on a smaller battery or a brighter display.

Accuracy is also a user experience problem. Rodriguez described the worst case plainly: the device shuts down while the user believes 16% charge remains. Many smartphones still struggle to give a trustworthy percentage, and the error tolerance shrinks further on devices with tiny cells.

Rodriguez also argues the problem is harder than it looks. Measuring the activity of the crystals and other materials inside a battery is difficult, and factors like dropping the device or temperature swings can shift the picture. So the chip can't be a simple sensor. It has to pair measurement with algorithms that adapt over time. That's why Overlord frames its product as a system rather than a component, and why the silicon and software have to be designed together.

Founders with a battery history

Nicholson holds a doctorate in microelectronics from Lancaster University and has led engineering and business teams at Texas Instruments and Fairchild. At Overlord he oversees platform development, along with the testing, qualification, and customer collaboration needed to turn semiconductor designs into products.

Rodriguez spent time on the executive and leadership team at Maxim Integrated, where he worked on product development, new product introductions, and strategic customer engagement. He later led global business development at Sevenhugs, which Qorvo acquired. He also helped define and launch a cloud-edge device at LatticeWork that won a CES Innovation Award.

His interest in batteries goes back further. A 1997 internship at PolyStor, which developed early lithium-ion prismatic cells for mobile phones, planted the seed. Nearly three decades later, he's arguing that the cell is still the constraint everyone designs around.

The two founders have decades of combined experience across semiconductors, consumer electronics, and venture-backed companies. They started Overlord about 16 months ago, and they're sampling chip designs with tier-one companies. The firm reached first silicon for its chip designs in July 2026.

Funding and the road to production

Overlord says it has raised $10 million to date. That includes a $7.25 million seed round and $2.75 million in non-dilutive strategic development funding. A later seed extension of $4.35 million was led by Band of Angels, with participation from Foothill Ventures, Unlock Pacific Ventures, NuFund Venture Group, Sand Hill Angels, Castle Fund, and returning strategic angel investors.

The team is about 15 full-time staff plus roughly 10 contractors, and it has filed for nearly 20 patents. Overlord's board includes Chris Allexandre, CEO of Navitas Semiconductor, and Dave Heacock, a former Texas Instruments senior vice president and general manager.

The company plans to raise another $15 million to $20 million and grow to about 60 people as it nears launch. The money will go toward validation, qualification, customer integration, and production infrastructure. With samples in customers' hands, the work now centers on product validation, firmware development, and production readiness.

Runtime, integration, and design constraints are what's pulling customers in, and their feedback will shape what the platform becomes. That's the part to watch. A chip like this lives or dies on whether design teams trust its numbers enough to shrink their safety margins. If Overlord can show that a cheap, updatable gauge lets a manufacturer drop that 30% buffer, the pitch stops being about a tiny chip and starts being about a bigger battery in a smaller frame.

Rodriguez's own summary is the most honest one: "We're a piece of technology that the customer will never see, but we will impact them on a daily basis. We're the hidden layer, and that's a great and sad place to be." He added, "You may never see it, but it could impact your life."