UK: BRITAIN'S CHEEKIEST CHIP SHOP - ARM. - ARM is a minuscule microchip designer from Nowheresville. It is taking on the giants of the industry and coming out ahead.

by Hashi Syedain.
Last Updated: 31 Aug 2010

ARM is a minuscule microchip designer from Nowheresville. It is taking on the giants of the industry and coming out ahead.

Who wants to be more commercial?' was the first question Robin Saxby put to his staff as they sat in the pub. It was spring 1991 and Saxby had just landed the job as managing director of ARM (Advanced RISC Machines), a tiny Cambridge-based designer of microchips, newly spun-off from Acorn Computers.

Three of the 12 assembled engineers, Jamie Urquhart, Mike Muller and Tudor Brown, put up their hands - and that was the first staffing problem sorted out. 'OK,' said Saxby, pointing in turn to the three engineers, 'you're sales director, you're marketing director and you're engineering director.'

Almost three years down the line, it is hard to fault the results of this somewhat unorthodox approach to management selection. ARM is taking on the giants of the microchip industry and, for the time being at least, is coming out ahead. In the past two years it has succeeded in signing up industry stalwarts such as GEC Plessey, Sharp and Texas Instruments to make and sell its chips. Apple, meanwhile, has chosen an ARM chip to power its hottest new launch - a hand-held personal communicator called Newton, due on the market in autumn this year. And US start-up 3DO is set to challenge the might of Nintendo and Sega in the video games market, with a state-of-the-art home multi-media gizmo powered by an ARM RISC chip.

For those uninitiated in the finer points of microchip engineering, RISC stands for Reduced Instruction Set Computer. RISC chips contrast with traditional CISC (Complex Instruction Set Computer) chips by concentrating on only the most common instructions. As Saxby puts it, there is an 80/20 rule, whereby most computers use just 20% of instructions 80% of the time. In practical terms, that means ordinary CISC microchips (like the Intel 486 in the average PC) are often bigger and more complex - ergo more expensive and slower - than they really need to be.

The market for RISC chips, explains Geoff Vincent, communications and electronics specialist with PA Consulting Group, is divided into two. The majority of RISC makers, such as Sun, Digital Equipment and others, have gone for high-performance, bells and whistles chips. ARM, meanwhile has established a lead in small, low-cost, low-power consumption chips. An ARM chip, for example, offers 50-80% of performance at a quarter of the size, an eighth of the power consumption and a fifth to a tenth of the price of its rivals. Other chip companies in the small end of the market, like AT and T with its Hobbit chip, have been unable to match that winning combination.

It means, says Vincent, that ARM chips are particularly attractive for the new generation of consumer electronics where price and power consumption are of the essence. Though ARM first developed chips for the schools' computer market, 'Now,' confesses Saxby, 'I'm afraid to say, that in terms of spending power, it's the games market that's driving it.' Kids today demand ever more value from multi-media gadgets that combine sound, video and sophisticated graphics. Nintendo, he points out, has profits of nearly £1 billion and fewer than 1,000 employees; 3DO, meanwhile, hopes to make its fortune by taking the games market one step further. Its home entertainment unit, due out before Christmas, is more powerful and more expensive than Nintendo or Sega, and will be billed as a family purchase, for the lounge rather than the children's bedroom.

But a bigger prize yet is the mega-high-volume, if rather less glamorous, embedded chip market - for the chips that control anything from power braking and fuel injection in cars, to the finely-honed browning technique of the 1990s toaster. The car industry, in particular, is a chipmaker's dream: an average car today contains 20 separate chips (a top-of-the-range BMW has nearly 50) and world production last year was over 33 million vehicles.

The story of ARM's ascent to microchip fame began with Acorn computers in the early 1980s. At that time Acorn, now 80%-owned by Italy's Olivetti, was riding high on the success of its BBC microcomputers. It had effectively cornered the UK schools market with over a million units sold, and was raring to develop a range of computers with much greater capacity. Chips on the market made by Intel or Motorola offered better performance than the Rockwell microprocessor which Acorn was using. But they were much more expensive - a serious problem for the cash-strapped schools sector. 'What they really wanted', Saxby says, 'was a tenfold increase in performance for no extra cost.' So Acorn, brimming with self-confidence and awash with money, decided to break into microchip design.

Steve Furber, Acorn's head of hardware design (now ICL professor of computing at Manchester University) and a colleague, Roger Wilson, set about hiring some sharp, young engineers and by the mid-1980s the first ARM chip was ready to hit the market. Acorn licensed US chip maker VLSI Technology to manufacture and sell them and the first ARM chips duly found their way into ICL and other computer peripherals as well as into the Acorn Archimedes.

But it was interest from Apple which really put ARM on the microchip map. Apple needed a chip for its Newton project, a pocket-sized notebook-cum-organiser-cum-fax that recognises handwriting and can communicate with other Newtons via infrared links. It was very interested in ARM's technology but wary of relying on a microchip controlled by a company that was ultimately a competitor. In 1990, together with VLSI, Apple persuaded Acorn to make ARM independent and took a 43% stake in the new spin-off for $3 million.

Saxby was headhunted as managing director and together with his 12 engineers - three of them elevated to management ranks - set up shop in a converted 18th century barn in the hamlet of Swaffham Bulbeck, a few miles beyond Cambridge. What Saxby brought to the young engineering team was management experience and marketing nous. Now 46, he has been in the business since 1968, first designing chips for TV sets, then moving into sales and management positions at Motorola. In 1984 he left to head a company making electronic security systems, before moving to a European microprocessor venture called European Silicon Structures (ES2). Part of the brief at ES2 was to turn round the company's loss-making US subsidiary - a task he achieved within a year through the sound management principles of increasing sales and cutting overheads. It was here that ARM's founders spotted Saxby - and poached him to be head of the new start-up venture.

Since he joined the company in 1990, Saxby has been the driving force behind ARM. In the electronics marketplace a careful business plan and, above all, good marketing are just as important as technological advantages. 'You've got to have a strategy and you've got to follow it,' Saxby declares. 'We're not trying to chase Intel on the desktop,' he states. 'We're targeting emerging markets like personal communications devices and multi-media where no one is yet established as a standard.' For a minuscule company based in Nowheresville, England, that is still a tall order - especially when you're competing against giants like Hitachi, Sun and AT and T. Right from the beginning, therefore, Saxby decided that selective licensing to high-profile companies was the only way forward. 'If you want to get your product accepted as a standard you have to win and be seen to be winning,' Saxby explains.

While his engineers worked on the first business goal - developing the new ARM 6 family for Apple's Newton in less than a year, Saxby turned his attention the next objective - finding a second licensee. By January 1992, he had signed up GEC Plessey, chosen 'because we needed a European partner. They had a 13-year history of profitability and they were physically close and convenient for working together.' Within six months of signing the deal, GEC Plessey was turning out ARM chips for the award-winning Acorn Series 310/320 computer range. It is an unusual British computer success story, and one of which Saxby is particularly proud. After some pretty ropy years, Acorn's profits leapt dramatically in 1992 to £1.3 million, thanks to increased sales of its new product range.

Events since then have moved at a dizzying pace. By the summer of 1992 ARM was ready to go to Japan - the heart of the consumer electronics industry. Saxby had had some dealings with the Japanese while working at ES2 and knew it was vital to go well prepared with properly translated documents, and the right connections. He hired a Japanese consultant, Masahiro Konishi, to set up seminars and meetings with the relevant decision-makers and then set off for Tokyo, 'to choose my Japanese partner'.

'Choose' may sound like an arrogant word for a handful of engineers in a barn near Cambridge, but ARM's reputation travelled fast. The Japanese eye for technological opportunity being what it is, most of the major semiconductor companies turned up to meet the Cambridge outfit. On top of this, at the very first seminar an excited Konishi came rushing up to Saxby to inform him that Takuro Isoda, president of Daiwa subsidiary Nippon Investment and Finance (NIF), wanted to take him out for dinner. 'This was obviously a great honour,' recalls Saxby. 'Everyone bowed 900 and he said, "I want to invest in your company".' Nine months later a deal was signed, giving NIF a 7% stake in ARM for £650,000.

As part of the process, Sadao Kunieda, head of NIF's international division, journeyed to Cambridge to make a presentation to the ARM board of directors. 'It was the only time he had ever had to convince a company to allow them to invest,' Saxby says. 'We have Apple and Acorn people on the holdings board, and it works because we get on well as people. We have a culture and that's what it's all about. We needed to know that these were people we could work with.' The strategic advantage of NIF's investment was not the capital it injected, but the doors it opened in the East Asian market.

While Saxby was shuttling between Tokyo and London to finalise NIF and sign up Sharp as an ARM licensee, his senior managers were targeting the US. Texas Instruments (TI) came to Cambridge in December 1992 and by May this year a deal was on the table. 'It was very exciting from a management point of view,' Saxby says. 'They did all the work, all I had to do was sign on the dotted line.' The tie-up with TI, the sixth largest semiconductor company in the world, is a further licensing coup for ARM because it offers an entree into the vast market for embedded control in the automotive industry. Strategically this market is especially important, not only for its size, but because product cycles are considerably longer than in consumer electronics. For the latter, says Saxby, you have to double performance every 18 months, whereas car chips are used for a period of years.

The American giant is now working with ARM to develop a chip for the auto industry that will combine several functions, such as braking, suspension and traction controls, on a single unit. It is a specially valuable endorsement of ARM, says Saxby, 'because they decided to take a licence instead of developing their own chip'. With all these big names under its belt, the ARM message has been spreading fast. Saxby is averaging an offer a week from companies wanting to be licensees - but he's playing hard to get. 'I'm only interested in people who can bring something new to the party. If someone comes to me who's just replicating markets already covered by our other licensees, I'm not interested,' he says.

ARM's rush of new deals has meant a wave of recruitment in Swaffham Bulbeck. The converted barn is at bursting point, with numbers trebled to nearly 40 engineers. There's little shelf space too for the rapidly multiplying Moet et Chandon bottles (18 at the last count), each label marked with the name of the chip it was drunk to christen.

Of course, ARM is tiny in the context of the global microchip industry. Licensees do all the manufacturing and most of the marketing, leaving Cambridge as the design nerve centre. 'We're running this company like a Silicon Valley start-up,' Saxby says, 'and in world terms we're very cost-effective.' ARM's own sales and marketing team concentrates on serving licensees and visiting important end users, as well as the company's small representative offices in the US and Japan.

All licensees and important end users are linked to ARM via electronic mail. Saxby himself spends so much time travelling he has dispensed with the notion of a desktop computer and relies instead on a small Apple PowerBook. Wherever he is, and whatever the time zone, he keeps up to date with developments in Cambridge by plugging his computer into a hotel phone line. 'I'm an expert in crocodile clips and taking phones apart,' he quips. 'I choose my hotels by the quality of their phone system.' Engineering capacity is multiplied, too, by teams from the various licensees who work with ARM to develop tailor-made chips for different applications. Put all these together, Saxby reckons, and there is probably more brain power going into ARM than into most of its competitors.

The next stage of Saxby's marketing plan is to co-ordinate as far as possible the marketing activities of all licensees. One initiative in the pipeline is a branded tag with the blue ARM logo which Saxby wants attached to all ARM-powered products. It is a profile-raising exercise which he hopes will turn the company into a household name, like Dolby stereo.

It is an irony for Saxby that, despite ARM's success, the company is still better known in Japan than at home in the UK. 'It's worrying for Britain. It's the smaller, dynamic companies that are really investing in technology,' he says. Of course, if everything goes as per Saxby's strategy, even Britons will learn of the exploits of ARM. At some stage too the company will be floated - which should make Saxby and his option-holding colleagues very rich men.

Since the mid-'80s the company has sold about one million chips. It should double that figure next year alone and could be selling tens of millions by the end of the decade. Saxby won't be drawn on predictions. 'People can easily get over optimistic in this business. The smart person says, "How's it going, how much can we really achieve?" and keeps overheads down till he's got the sales. That's why they hired me - because I'm 46 and I'm battle scarred.'.

Find this article useful?

Get more great articles like this in your inbox every lunchtime