The Cambridge Spinoff That Ended Up Inside Every Smartphone on Earth: Arm's Unlikely Story

The Cambridge Garage Spinoff That Ended Up Inside Every Smartphone on Earth

If you're reading this on a phone, there's a near-certainty that the chip inside it was designed around technology from a single British company: Arm. Not Apple, not Qualcomm, not Samsung — Arm doesn't make a single chip itself. It designs the architecture that practically everyone else licenses, and the story of how a 12-person spinoff from a struggling British computer maker ended up underpinning the entire mobile industry is one of the stranger footnotes in tech history.

Born From a Dying British PC Maker

Arm's roots go back to Acorn Computers, a Cambridge, UK firm best known in the 1980s for the BBC Micro, a computer that taught a generation of British schoolchildren to code. By the mid-1980s, Acorn's engineers — including Sophie Wilson and Steve Furber — were frustrated that no commercial chip could do what they needed for their next machine. So they designed their own: the Acorn RISC Machine, first working silicon in 1985, using a simplified "reduced instruction set" (RISC) approach that traded raw complexity for efficiency and low power draw.

That efficiency is what later made the design so valuable, but at the time it was just a clever engineering choice for a cash-strapped company that couldn't afford Intel-scale chip budgets.

The Apple Connection Nobody Expects

Here's the part that surprises people: Arm, as an independent company, exists because of Apple. In November 1990, Acorn spun its chip design team out into a new joint venture called Advanced RISC Machines Ltd — backed by Acorn, chip manufacturer VLSI Technology, and Apple, which needed a low-power processor for an ambitious new handheld project called the Newton. Apple's investment (reported at around $3 million) and Acorn's dozen engineers became the founding nucleus of the company. Apple reportedly insisted the new venture not carry its rival Acorn's name, which is part of why it became "Advanced RISC Machines" rather than keeping the Acorn branding.

The Newton itself flopped commercially, but the chip architecture built to power it didn't. Robin Saxby, recruited as the new company's first CEO, pushed it toward an unusual business model: instead of manufacturing chips, Arm would simply license its instruction-set designs to anyone willing to pay — Texas Instruments, Samsung, and eventually nearly every mobile chipmaker on the planet.

Why Licensing, Not Manufacturing, Won

That decision is the whole story. Building and operating a chip fab costs billions of dollars and takes years. Arm instead sold blueprints: companies like Qualcomm, Apple, Samsung, and MediaTek could take an Arm core design, customize it, and have it manufactured by foundries such as TSMC. It meant Arm never had to bet its own capital on silicon fabrication, and it meant chipmakers got a mature, power-efficient design without the enormous cost of inventing one from scratch.

This asset-light model turned out to be perfectly timed for the smartphone era. Battery-powered devices needed exactly the kind of power efficiency Arm's RISC architecture had been built around out of necessity in the 1980s. By the 2010s, Arm-based designs were running the overwhelming majority of the world's smartphones, and today Arm says its architecture ships in well over 99% of them, alongside a growing share of laptops, servers, and data center chips.

SoftBank's Big Bet — and the Deal That Collapsed

Arm's independence didn't last forever. In 2016, Japanese conglomerate SoftBank, led by Masayoshi Son, acquired Arm Holdings for roughly $32 billion — a bet that chip architecture, not just the chips themselves, would become increasingly central to computing's future. It was one of the largest-ever acquisitions of a UK tech company.

Then, in 2020, SoftBank tried to sell Arm again — this time to Nvidia, for a deal that eventually ballooned toward $40 billion as Nvidia's stock price climbed. That deal became one of the most closely watched antitrust fights in chip history. Regulators in the US, UK, and EU all raised competition concerns, worried that Nvidia owning the architecture nearly every rival chipmaker depends on would give it outsized leverage over the entire industry. In February 2022, SoftBank abandoned the acquisition, forfeiting a $1.25 billion breakup fee Nvidia had already put down.

Going Public, Again

With the Nvidia sale dead, SoftBank took Arm public instead. The company priced its Nasdaq IPO at $51 a share in September 2023 — the top of its announced range — raising about $4.87 billion and valuing the company at roughly $54.5 billion. Shares jumped nearly 25% on their first day of trading, pushing Arm's market value toward $65 billion. It was the largest US tech IPO in almost two years, and SoftBank retained the vast majority of the company — holding over 90% of shares after the offering.

Why This Still Matters

Arm's story matters beyond chip-industry trivia because it illustrates something about how modern technology actually gets built: the company whose intellectual property is in almost every pocket on the planet is barely a household name, because it never had to be. It never sold a single consumer product. Instead, it bet on being the invisible layer everyone else builds on top of — a strategy now being echoed by other IP licensors in AI accelerators, networking chips, and custom silicon more broadly.

As more companies, from Apple to Amazon to Google, design their own custom chips, Arm's licensing model is more relevant than ever. Those firms don't start from scratch; they still license core Arm architecture and build their own customizations on top, a pattern that traces directly back to a scrappy, underfunded spinoff trying to solve a very specific problem for a doomed 1990s pocket computer.