The Intel spinout hiding in your dashboard

Today, I'm digging into Mobileye Global (MBLY).

They make the cameras, chips, and software that power modern car safety features. If your vehicle does lane-keeping, automatic emergency braking, or adaptive cruise control, there's a decent chance a little Mobileye "EyeQ" chip is doing the looking. They've sold that chip into more than 200 million cars across roughly 50 automakers.

I've never driven a car with full self-driving, but I'm a huge fan of adaptive cruise control. It has made driving on longer trips considerably more enjoyable for me. Specifically, I can listen to podcasts and audiobooks, which are things I love to do. I find that I am able to absorb more of the content, so I'm a big fan of what Mobileye has been able to do!

A few things that stood out to me:

• They own the “middle” of self-driving. Instead of chasing full-autonomy like Waymo and Tesla, Mobileye supplies the camera-based ADAS already in mass-market cars, around two-thirds of the global market. Their brand isn't well-known to consumers, but it's a staple supplier to auto makers.

• They offer a lesson in how gap accounting can get messy when trying to understand the true financial performance of a company. GAAP shows big operating losses but they produced $523M of free cash flow in 2025 and have $1.8B of net cash (meaning no debt). The losses are mostly non-cash amortization from Intel's purchase accounting plus a one-time $2.7B goodwill write-down in 2024.

• They have a real data moat. Every EyeQ car can crowdsource road data through Mobileye's REM mapping. Two hundred million cars feeding a driving dataset is an asset that is hard to replicate.

But it's not all clean and perfect:

• China is roughly 23% of where product ships, and a homegrown rival (Horizon Robotics) is gaining traction.

• Surprisingly, they have had nearly flat revenue over the last few years.

• The robotaxi story is still mostly a promise. The Lyft launch already slipped from 2026 to 2027.

• Fear of customer insourcing. Tesla already proved an automaker can build its own perception stack. The worry is that more large OEMs, especially in China, decide ADAS is core and bring it in-house

The last three bullets are a big reason why the stock is down over 50% over the last year.

The key takeaway here for me is that they are in a pretty powerful position to get designed into the next generation of more fully autonomous cars. They have proven relationships with dozens of automakers, a ton of data to be used to continually improve their offerings, and they have the natural product ladder for companies to step up into as autonomy increases.

The big question mark to me is whether Tesla is going to dominate self-driving, or if there will be many automakers in the game. If there are, Mobileye is incredibly well positioned. It's going to be fun to watch it play out

With that, I'll see you tomorrow!

Nick

TL;DR

• Mobileye builds the cameras, EyeQ chips, and software behind driver-assistance and self-driving features, sold to ~50 automakers and shipped into 200M+ cars.

• Revenue is overwhelmingly base ADAS today, with SuperVision, Chauffeur, and robotaxi "Drive" as the growth bets.

• Performance has been bumpy: a 2024 China inventory correction crushed revenue, which recovered to $1.9B in 2025.

• Unusual financial profile: GAAP operating losses, but ~48% GAAP gross margin (closer to 70% adjusted), $523M of free cash flow, and $1.8B of net cash with no debt.

• The stock has been a near-term wreck: down ~56% over the past year versus Aptiv at -8% and Ambarella at -1%, though up ~13% over the last three months.

• Operator takeaway: an installed base nobody can replicate quickly, plus a step-by-step product ladder, keeps customers even when headlines say your category is about to be upended.

The 30,000-Foot View

Mobileye sells "eyes and a brain" for cars. The core product is the EyeQ system-on-chip, paired with computer-vision software, that an automaker designs into a vehicle platform to power features like automatic emergency braking, lane centering, and adaptive cruise. It's sold mostly through the big Tier-1 auto suppliers, who integrate it and ship it to the carmaker. Base ADAS is where almost all the money comes from today, and Mobileye has roughly two-thirds of the global market for camera-based driver assistance.

What makes it interesting is the upgrade ladder. Mobileye productized the entire path to autonomy: basic ADAS at the bottom, then SuperVision (eyes-on, hands-off driving), then Chauffeur (eyes-off L3), then Drive (the full robotaxi stack). Same silicon family, same automaker relationship, climbing the rungs over time. On top sits REM, a crowdsourced mapping system where EyeQ cars send back road data that sharpens Mobileye's high-definition maps, plus their own imaging radar to cut reliance on third-party sensors.

It's a design-in business: getting onto a platform takes years of validation, and once you're in, you stay for its life. That creates real switching costs, lumpy program-driven revenue, and means today's design wins are tomorrow's revenue.

Revenue mix by ship-to region (FY2025, approximate): China ~23%, US ~22%, Germany ~16%, South Korea ~10%, UK ~6%, rest of Europe and world ~23%. (That reflects where Tier-1 manufacturers take delivery, not where the end driver lives.)

Key Stats

• Market cap: ~$6.4B

• FY2025 revenue: ~$1.9B

• GAAP gross margin: ~48% (adjusted gross margin is closer to 70%)

• Employees: ~4,200

• Industry: ADAS and autonomous-driving technology

• 1Y total return: about -56%

Company History

• 1999: Amnon Shashua, a Hebrew University computer-vision professor, co-founds Mobileye with Ziv Aviram in Jerusalem, betting a single camera plus smart software could read the road.

• 2007: The first EyeQ chip ships, kicking off the franchise that still drives the business.

• 2014: Mobileye IPOs on the NYSE, then the largest Israeli IPO in US history.

• 2017: Intel acquires Mobileye for about $15.3B as its autonomous-driving arm.

• 2022: Intel spins Mobileye back out via a Nasdaq IPO in October, keeping majority control.

• 2023: Revenue peaks around $2.1B as automaker programs ramp.

• 2024: A brutal year. A Tier-1 inventory correction gutted revenue to ~$1.7B, the company took a ~$2.7B non-cash goodwill impairment, and Intel signaled it would cut its stake.

• 2025: Revenue recovers to ~$1.9B with $523M of free cash flow, plus SuperVision and Surround ADAS design wins (including Mahindra) and progress on the Volkswagen/MOIA robotaxi program.

• 2026 (YTD): Q1 brings an updated outlook and a new $250M buyback. Volkswagen stays on track for US driverless deployment; the Lyft robotaxi launch slips to 2027.

Show Me the Money

Standout financial features:

• Revenue is cyclical and program-driven, not a smooth subscription line. It went $1.9B, $2.1B, $1.7B, $1.9B over four years, and the 2024 dip was an inventory correction, not a demand collapse.

• The GAAP operating losses are mostly optical. The 2024 loss of $3.2B includes a $2.7B goodwill impairment, and every year carries ~$500M of amortization from Intel's 2017 purchase accounting. Strip those out and this is a roughly breakeven-to-profitable, cash-generating business.

• Cash is the real tell: $523M of free cash flow in 2025 and $1.8B of net cash with no debt, a pile that keeps growing.

• That same amortization runs through cost of goods, so the ~48% GAAP gross margin understates the real economics, which look closer to a 70% adjusted margin.

Financial Data

(Net Debt is negative because Mobileye carries net cash, not debt. The FY2024 operating loss is dominated by a one-time $2.7B goodwill impairment.)

Stock Performance

(5Y and 10Y are unavailable because Mobileye only re-IPO'd in October 2022, so it has less than four years of public history in this incarnation.)

1-year head-to-head versus peers:

Mobileye badly lagged both comps over the past year. The ~13% bounce over the last three months suggests some of the panic (China destocking, the Intel overhang) is draining out, but a 56% one-year drawdown against roughly flat peers tells you the market is still skeptical the growth bets pay off on schedule.

The N.O.O.B. Nine: Competitive Powers

The Nerd Out on Business Nine is made up of Hamliton Helmer's famous "7 Powers" of competitive advantage (Scale Economies, Network Economies, Counter-Positioning, Switching Costs, Branding, Cornered Resource, and Process Power) combined with two of my own (Data Flywheel and Distribution Advantage).

Average Score: 3.2/5 - The moat is scale, design-in switching costs, and a genuine data flywheel, undercut by rising competition and the risk that big customers eventually build perception in-house.

Memorable Marketing

Mobileye does almost no consumer marketing, which fits a company whose product hides inside someone else's car. Its real marketing is engineering credibility aimed at automakers, regulators, and investors, and the megaphone is Amnon Shashua.

Notable tactics:

• The CES keynote spectacle (annual): Shashua's CES presentations are the company's biggest stage for robotaxi updates, new chips, and live autonomous-drive footage. Product launch as thought leadership.

• Open-sourcing RSS (2017 onward): Mobileye published its "Responsibility-Sensitive Safety" model, a formal framework for safe autonomous driving, and pushed it toward industry standards. Giving away the rulebook shapes the rules you'll be judged by.

• Public autonomous-drive demos: Livestreamed drives through dense cities (Jerusalem, Detroit, New York) prove capability more credibly than any ad.

• The product-name ladder (EyeQ, SuperVision, Chauffeur, Drive): The naming turns an abstract autonomy spectrum into a clear menu an automaker can shop from.

Tactical takeaways:

1. If your buyer is an engineer, sell with demos and standards, not slogans.

2. Give away a framework (a safety model, a spec) to anchor how the market evaluates you.

3. Make your founder the brand when the founder is a credible technical authority.

4. Name your product tiers so customers see the upgrade path before they need it.

AI Uses & Opportunities

Current exposure: Mobileye is fundamentally an AI company. The EyeQ chip exists to run deep-learning perception, and the value proposition is computer vision plus driving-policy models that turn camera pixels into driving decisions. It already monetizes that directly: better models mean more capable ADAS and higher-value design wins as automakers climb from basic features to SuperVision.

Future opportunities:

• Licensing driving "foundation models" and the full self-driving stack to automakers who don't want to build it.

• Selling or licensing the REM map and aggregated driving data as a standalone asset.

• AI-driven simulation to validate autonomy faster and cheaper than real-world miles alone.

• Robotaxi-as-a-service economics, where the Drive software stack earns per-mile or per-vehicle revenue instead of a one-time chip sale.

Bumps in the Road

• China concentration and competition. China is ~23% of where product ships, and local champion Horizon Robotics targets that ADAS market with cheaper, "good enough" options.

• The Intel overhang. Intel still controls ~80% of the stock and wants out. Every secondary offering adds fresh share supply, and controlled-company governance limits minority shareholders' say.

• Customer insourcing. Tesla proved an automaker can build perception in-house. If more large OEMs follow, Mobileye loses the design win and the data that comes with it.

• Lumpy, program-driven revenue. The 2024 inventory correction showed how fast the top line swings when Tier-1 customers adjust stock, even without a demand problem.

• Autonomy timelines that keep slipping. The robotaxi and L3 story is real but slow, and the market is paying for proof, not promises.

Your Swipe File

• An upgrade ladder is powerful. EyeQ to SuperVision to Chauffeur to Drive lets one customer relationship grow with the roadmap.

• Turn your installed base into a data moat. Two hundred million cars sending back road data is a hard-to-replicate asset. Does your business offer any unique data sources that can drive self-improvement or offer secondary monetization opportunities?

• When looking at public companies, it's important to understand the impact of non-cash charges. Look deeper than the income statement when analyzing financial statements.