Founding of ARM and the Acorn RISC Machine

You might be surprised to learn that ARM's story starts with Acorn Computers, a small British firm that developed its first RISC processor in 1983 using just a handful of engineers. That original chip actually outperformed Intel's 80386 while using fewer transistors. Acorn eventually spun ARM off as its own company in 1990, launching a licensing model that shipped over 10 billion processors by 2008. There's much more to this fascinating origin story ahead.

Key Takeaways

• Acorn Computers founded ARM as a joint venture specifically to develop a RISC processor, later spinning it off as a separate company in 1990.
• The original ARM processor outperformed Intel's 80386 chip while using significantly fewer transistors, demonstrating remarkable efficiency.
• ARM's unique licensing model, rather than manufacturing chips itself, became the foundation of its global market dominance.
• By 2008, ARM's licensing model had enabled over 10 billion ARM-based processors to be shipped worldwide.
• ARM's focus on royalties from licensed chip production funded the development of successive, more advanced processor generations.

What Sparked Acorn's RISC Obsession in 1983?

In 1983, Acorn wasn't just building computers — they were chasing a market that IBM had already proven existed. IBM's PC launch showed you could command premium pricing in business computing, and Acorn recognized that business market opportunity through their ABC development plan.

But building the ABC exposed a serious problem: existing processor limitations made the goal nearly impossible. The Motorola 68000 couldn't meet the BBC Micro's communication protocol requirements, and the National Semiconductor 32016 actually performed worse than Acorn's own 4 MHz 6502. That's embarrassing for a machine meant to compete professionally.

Meanwhile, Apple Lisa was proving that graphical interfaces were the future — a future that no available processor could adequately support. Acorn didn't just hit a wall; they hit every wall simultaneously. Acorn had already demonstrated its engineering ambition years earlier, having dominated the educational computer market with the BBC Micro throughout the 1980s.

The solution Acorn pursued drew directly from academic research, as they adopted the Berkeley RISC concept to optimize for a limited set of efficient instructions that existing commercial processors simply weren't designed to execute.

How Did the ARM1 Chip Come to Life in a Single Day?

When Acorn's engineers finally powered up their first ARM1 silicon on 26 April 1985, the chip worked — completely, correctly, on the very first attempt. That single day chip development milestone was extraordinary, especially given the early prototyping challenges the team faced.

The initial test board even had a broken power connection, yet the chip still functioned by drawing enough current through its own leakage. You'd expect months of debugging after receiving first silicon, but ARM1 demanded none of that.

Steve Furber had completed the tape-out in Munich just months earlier, and the result was a chip containing fewer than 25,000 transistors running at 6 MHz, consuming only 0.1 watts. That first day of testing immediately enabled development of support chips VIDC, IOC, and MEMC. The ARM1 was subsequently used as a second processor for the BBC Micro, helping the team develop simulation software and accelerate CAD work for the ARM2.

The ARM1's data path occupied the bottom half of the chip, operating on 32 bits at a time using horizontal data buses to process information efficiently across the silicon.

The Archimedes and Britain's First ARM-Powered Home Computer

Two years after ARM1's successful debut, Acorn launched the Archimedes in 1987, making it the world's first home computer series to run a RISC processor.

The Archimedes hardware specifications centered on the ARM2, an 8 MHz, 32-bit chip with roughly 27,000 transistors delivering 4 MIPS. That's impressive considering it outpaced the Amiga 500's 68000 by seven times and processed three times more instructions per clock cycle than the Intel 80386.

Acorn later upgraded to the ARM3 at 25 MHz in 1990, adding on-chip cache for even greater speed. Despite strong Archimedes market performance in Britain, where it built on the BBC Micro's legacy, the system struggled to gain traction internationally, limiting its broader commercial impact. The Archimedes A3020, released in 1992, helped narrow the performance gap with the then-new Amiga A1200.

The Archimedes also made a notable mark in gaming, becoming one of the first home computers to feature a high number of 3D games, with smooth filled polygon titles running at up to 25 FPS on the base 8 MHz hardware.

How Apple's Newton Problem Gave ARM Its Defining Moment

By 1987, Apple had kicked off the Newton project, hunting for a processor that could power a handheld PDA without draining its battery in hours. VLSI Technology pointed Apple toward Acorn's ARM chip, and by fall 1990, Apple, Acorn, and VLSI had negotiated a joint venture, forming Advanced RISC Machines Limited. Apple's $3 million secured a 43% stake, and the partnership produced the ARM610 that powered the Newton MessagePad's 1993 launch.

Newton's commercial failure forced ARM's CEO Robin Saxby to rethink everything. He couldn't let ARM's survival depend on a single product. Newton's influence on ARM's business model pushed Saxby to introduce ARM's licensing strategy, charging upfront fees plus royalties. That pivot attracted Texas Instruments and Nokia, transforming ARM into the global processor standard you recognize today. ARM's RISC design had originally caught Apple's attention in the late 1980s, making the eventual partnership a natural extension of that early interest.

When ARM IPO'd in 1998, Apple sold its stake for $800 million, providing a critical cash infusion at a time when the company desperately needed funds to reset and return to profitability.

The Barn Where ARM Ltd Was Born

On November 27, 1990, ARM Ltd officially launched inside a converted 17th-century barn nestled in Swaffham Bulbeck, a picturesque village just outside Cambridge. Despite its humble origins, the barn design was surprisingly functional, housing 12 engineers transferred from Acorn alongside Apple's £1.5M cash investment and VLSI Technology's tools and funding.

They'd no patents, no independent customers, and a product that wasn't yet market-ready. Still, they secured a second licensee by 1992 and turned their first profit in 1993. By March 1994, the company had outgrown the barn entirely, relocating to Fulbourn Road as headcount surged past 70.

You might think a barn would signal a shaky start, but the team hit the ground running despite serious initial product challenges. The ARM7TDMI, introduced in 1994, became a landmark achievement, representing the first widely adopted Arm processor and powering mobile devices like the Nokia 6110. The company was structured as a joint venture between Acorn, Apple, and VLSI Technology, a tripartite arrangement that gave ARM its foundational resources and credibility from day one.

Why Renaming the Chip Changed Everything

When Acorn and Apple spun off the ARM architecture into a joint venture with VLSI Technology in November 1990, they faced an immediate branding problem: "Acorn RISC Machine" tied the chip's identity to a British home computer company that Apple had no interest in being associated with.

Renaming it "Advanced RISC Machines" stripped away that baggage and repositioned the processor as a neutral, licensable technology. That subtle shift opened up a competitive advantage no single product company could've achieved alone. Instead of selling chips, ARM sold intellectual property, letting manufacturers worldwide build on the architecture.

As Acorn's original machines faded into obsolescence, the new name meant nothing historically — and everything commercially. That clean slate became the foundation for ARM's eventual market dominance across mobile and embedded computing. The original ARM processor had already proven its potential, outperforming the Intel 80386 while using less than one-tenth the transistors. By 2008, the company's licensing model had enabled over 10 billion ARM-based processors to be shipped worldwide.

The Royalty Deal That Turned ARM Into a Global Empire

The joint venture's true genius wasn't the chip itself — it was how ARM chose to monetize it. Instead of manufacturing chips, ARM collected upfront licensing fees plus royalties on every chip shipped. That licensing flexibility let any company integrate ARM architecture without ARM touching a factory floor.

The results were staggering. By 2015, manufacturers were producing 15 billion ARM-based microprocessors annually. Every chip meant another royalty payment flowing back to ARM. That revenue scalability turned a modest joint venture into a global force, funding successive architecture generations without ballooning production costs.

What started as upfront cash from Apple and tools from VLSI ultimately compounded into one of technology's most valuable intellectual property empires. ARM's reach extended so broadly that ARM-based chips are found in the majority of the world's smartphones and tablets. You can trace the endpoint directly to 2016, when SoftBank acquired ARM for £24 billion. In 2023, ARM further cemented its status as a public company through a $4.87 billion IPO on the NASDAQ exchange.