On March 24, 2014, Intel announced it had acquired Basis Science, a four-year-old startup that had shipped exactly one product. The press release was short. The price was not disclosed. By the time the recall paperwork hit the Consumer Product Safety Commission’s desk two and a half years later, the deal had become a small case study in why semiconductor companies are bad at buying consumer hardware businesses.
What follows is the corporate story of the Basis Peak, separated from the product story. It is also an answer to a question we get more than any other: how did Intel spend, by the high estimate, $150 million, ship a watch that overheated, and then quietly leave the category?
The deal, as it was reported
The first credible reporting came from Liz Gannes at Re/code on March 24, 2014, with TechCrunch confirming the same day. Both outlets cited the price as “around $100 million.” Neither company commented on the number. Within a few weeks, Forbes and the wearables-industry tracker Wearable Technologies ran higher estimates, and $150 million was the figure that stuck in later secondary sources. Intel has never confirmed a price publicly, and Basis’s founders have not, either.
For an acquisition of that size, the silence is unusual but not unprecedented. Intel disclosed it as immaterial, which is to say it did not break out the purchase price in its 10-Q. That alone is a useful signal: if it had been a $300 million deal, the SEC filings would have flagged it.
What was reported, with more confidence, is what Intel got:
- Basis Science, founded in 2010 in San Francisco by Jef Holove and a team that came out of Stanford and the wearables-research community.
- The Basis B1, a round-faced wrist tracker that had shipped in late 2012 and had become a darling of the digital-health press for fusing optical heart rate, galvanic skin response, skin temperature, and a 3-axis accelerometer in a consumer device.
- An engineering team of roughly 50 people, most of them in San Francisco.
- Funding history that included Norwest Venture Partners, Mayfield Fund, Doll Capital Management (DCM), and the Stanford-StartX Fund. Total disclosed venture funding before the acquisition was around $32 million across three rounds.
The strategic rationale, in Intel’s framing, was wearables. Two weeks before the Basis announcement, Intel had launched its “Make It Wearable” challenge, a $1.3 million developer competition aimed at proving the company had a story in the category. The Basis deal was the proof point: an actual product, an actual team, an actual install base of customers wearing Intel-owned hardware every day.
Why Intel wanted in
It helps to remember what 2014 looked like from Santa Clara.
The PC business was shrinking. Mobile had effectively been lost to ARM. Brian Krzanich had taken over as CEO in May 2013 with an explicit mandate to find new growth categories, and the Internet of Things (wearables, smart home, industrial sensors) was the bet. Intel had set up the New Devices Group (NDG) under Mike Bell, a former Apple and Palm executive, in early 2013 specifically to chase this. The group’s job was to produce reference designs and finished products that demonstrated Intel silicon (specifically the new Quark and Edison platforms) in form factors that PC chips would never reach.
Basis Science fit. It was a finished product with a credible brand, a defensible sensor stack, and a team that had already proven it could ship hardware. The B1 used a custom ASIC plus an off-the-shelf MCU; the obvious next step was to spin a Basis-on-Intel-silicon device, which is roughly what the Peak turned out to be.
There was also a defensive dimension. Google bought Nest in January 2014 for $3.2 billion. Apple was rumored to be working on what became the Apple Watch (announced September 2014, shipped April 2015). Samsung had already shipped the original Gear in 2013. Intel did not want to be the only large-cap tech company without a wearables story when the analyst calls came around.
What happened after the close
Basis CEO Jef Holove stayed on for about a year and then left. The team moved into Intel’s San Francisco office and was absorbed into NDG. Mike Bell ran the group; the Basis engineering leadership reported up through him.
Eight months after the deal closed, the Basis Peak shipped — November 2014, at a $199 retail price. The Peak was the redesign Basis had been working on before the acquisition: square OLED, more aggressive sport-watch shape, a notification model that tried to compete with the Apple Watch’s sketch sheet. Intel marketing put it in the “smart fitness watch” category, splitting the difference between Fitbit and Apple.
For a year, things looked normal. The Peak got decent reviews. Sales were not disclosed, but the unit count we now know, roughly 200,000 affected by the eventual recall, implies a modest but real install base for a first-generation product. There was a second SKU, the Peak Titanium, that shipped in 2015 with a metal band and a steeper price.
Then, in November 2015, Intel pushed a firmware update meant to improve the optical heart rate sensor’s behavior during high-intensity exercise. Within weeks, the Basis support forums filled with reports of the watch heating up and, in some cases, blistering the wrist. Intel issued a software fix in March 2016. It did not solve the problem. In June 2016, the company switched to voluntary refunds, and on August 4, 2016, the formal CPSC recall notice (16-235) was published. We cover the technical and regulatory side in the recall write-up and the product autopsy in why the Basis Peak failed.
The cloud went dark on December 31, 2016. The apps came off the App Store and Google Play around the same time. The Peak became, effectively, a paperweight with a working OLED. We have a separate retrospective review of the Peak for anyone who wants the wearer’s perspective.
The NDG wind-down
Four months after the cloud shutdown, in April 2017, Intel told staff it was restructuring NDG. The trade press reported roughly 140 layoffs and a reorientation away from consumer wearables toward “augmented reality and other emerging technologies.” Mike Bell left Intel that year. The Recon Jet smart-glasses team, acquired separately in 2015, was wound down on a similar timeline. By 2018 there was nothing left of the consumer wearables operation that the Basis deal had been the centerpiece of.
Intel never gave a public post-mortem. The closest thing to an official statement was a Krzanich quote in 2017 about Intel needing to focus on data-center and AI silicon. That is true on its own merits. It is also the kind of statement a CEO makes when a strategic bet has not worked.
How this compares to peer-era hardware acquisitions
Intel was not the only large company to misjudge consumer hardware in this window. The useful comparison set is the 2013–2017 wave of platform companies buying connected-device startups.
| Acquirer | Target | Year | Reported price | Status by 2020 |
|---|---|---|---|---|
| Nest | Jan 2014 | $3.2B | Re-absorbed into Google hardware after Tony Fadell’s 2016 exit; brand survives, original ambitions sharply curtailed | |
| Intel | Basis Science | Mar 2014 | about $100M–$150M | Discontinued 2016, team disbanded by 2018 |
| Under Armour | MapMyFitness / MyFitnessPal / Endomondo | 2013–2015 | about $710M combined | MyFitnessPal sold to Francisco Partners in 2020 for $345M, a write-down |
| Fitbit | Pebble (assets) | Dec 2016 | about $23M (software only) | Hardware line shut down; key staff went to Google |
| Fitbit | 2021 close | $2.1B | Brand still active but much narrower, Sense 2 effectively the last new platform |
The pattern is consistent. Companies whose primary business is software, services, or silicon repeatedly overestimated how much of that strength would transfer to a wrist or thermostat. Hardware has different timelines, different margins, and a different set of failure modes — the Basis Peak recall is the cleanest illustration in the set, but it is not unique.
The closer analog to Intel/Basis is probably Google/Nest. Both were paid for in cash by acquirers with deep pockets. Both had charismatic founder-CEOs who left within two years. Both saw their roadmaps absorbed into the acquirer’s internal politics and slowed accordingly. The difference is that Nest had enough installed base and brand equity to survive a re-org; Basis did not.
What Intel got wrong, in our view
We have written about the product side of this elsewhere. The corporate-side mistakes were three, and they compound.
First, NDG was structured as a skunkworks. Mike Bell reported high in the org but his group sat outside the main P&L. That meant fast decisions early on, but it also meant that when the Peak began to wobble, there was nobody in the consumer-hardware operations chain who could be held accountable for the recall in the way that, say, a Fitbit COO would have been. The team that built the watch was not the team that handled the CPSC filing, and the team that handled the CPSC filing was not the team that ran the support forums.
Second, Intel bought a hardware company without buying a hardware supply chain. Basis had used contract manufacturers; Intel did not absorb those relationships in a way that gave it visibility into the battery supplier. The recall traced back to a charging-cycle interaction in specific units; a more vertically integrated acquirer might have caught it earlier in QA, or at least sourced replacements faster.
Third, the strategic frame was wrong. The Basis deal was a wearables play, but Intel’s actual advantage was in silicon. The version of this story where the deal works is the one where Intel sells the Curie SoC to ten Fitbit-shaped customers and Basis is the reference design. Instead, Intel competed with its potential customers. Fitbit, the obvious partner, never shipped an Intel chip. By 2017 there was no reference-design customer base to fall back on, and the consumer product had been recalled.
What outlived it
A few things from the Basis era are still detectable in the modern wearables market.
The multi-sensor fusion approach (combining optical heart rate, skin temperature, and electrodermal activity) is now standard. Oura built its brand on it, the Apple Watch Series 8 added skin temperature in 2022, and the Fitbit Sense had electrodermal activity from launch. None of these cited Basis directly, but the engineering vocabulary they use is the one Basis Science was speaking in 2013.
The other survival is institutional. A meaningful number of the engineers who worked on the Peak ended up at Apple, Google, and Fitbit between 2016 and 2018. The B1’s lead biomedical engineer went on to Verily. The Peak’s optical sensor team partly reformed inside Google’s hardware org. The product died; the people did not.
The acquisition itself, viewed from 2026, looks less like a strategic mistake and more like the kind of expensive education that large companies sometimes pay for. Intel spent somewhere between $100M and $150M, plus three years of headcount, to learn that consumer hardware is not silicon licensing with a strap on it. Whether that lesson was worth the price is a question for Krzanich’s successors, who have, to their credit, not repeated the experiment.