The ARM story: Success, no stress and lots of happiness

Cambridge-based ARM is heralded as a chips success story in these tricky times. Here Kate Hanaghan looks at the spanners that might just end up in ARM's works.

The ARM story reads like a fairytale compared to the rest of the chip industry. Its strategy of diversity in adversity seems to be paying off and is reflected in it achieving an enviable 46 per cent increase on last year's profits.

Over the years it has quite cleverly spread itself across eight industry sectors, which means it's not reliant upon income from one area alone.

ARM produces generic technologies that can be tweaked and applied to numerous markets. And the benefits are twofold. First, the vast bulk of R&D can be common: invest once but get multiple uses and more importantly multiple royalties. This also means the company is perfectly poised to pounce on any emerging markets or technologies without any significant overhaul of its designs. Clever.

ARM's most prudent move was to gain huge first mover advantage in the wireless market. Worldwide, the company has chip designs in over three quarters of all mobile phones.

And while the rest of us are worrying about the lack of 3G products, ARM is more than confident that it will continue to reap the rewards. In the meantime, there is plenty of 2/2.5G action to keep things ticking over nicely.

However, even the best parties have to end sooner or later and looking ahead there could be trouble.

Jim Tully, analyst at Gartner, said: "While there is so much talk about certain architectures such as Intel, ARM and MIPS, who actually cares about that? When the area on the processor that the chip sits on becomes so tiny do we care about what it's doing? Is it possible that we'll see middleware emerge that will mask the chip?"

ARM's design engineering process is not 'synthesisable'. This means the designs are not fed into a computer program that can spit out the designs. This process would probably make ARM more efficient and would enable it to churn out designs faster and with greater flexibility.

Conversely, who wants to be bound by the mindless constraints of a computer program?

And recently ARM has begun to find that life at the top is hard. There are always pretenders to the thrown snapping viciously at your heels.

Alex Stuart, director of semiconductor research at IDC, explains there are young upstarts looking to take a swipe at what is potentially a very lucrative market. He said: "They are trying to do to ARM what AMD did to Intel."

One such company is picaTurbo. In June, the company - which produces synthesisable designs for RISC microprocessor cores - won a ruling supporting its claim that its designs do not infringe upon certain patents owned by ARM. No so good news for the Cambridge company.

Paradoxically, ARM's own customers might also pose a threat. The company has a range of licensee customers from colossal household names such as Intel to small and completely unknown companies. This set-up could create issues for the company in the future.

Shane Rau, research analyst at IDC, argues some of the very large licensees "could take the ARM design and effectively hijack the architecture". This could happen, he suggests, if the smaller licensees get squeezed out by the bigger boys in terms of contributing their designs to the ARM architecture.

But just maybe ARM has realised this. The company's PrimeXsys platform comprises a bunch of tools for designing applications with an emphasis on time and cost efficiency. Although ARM has not presented it as such, this platform might be an intentional method for bolstering the contributions of smaller licensees.

The going might still be good at the moment for ARM but things could change. If it's as agile as its management would have us believe, it might just have the capability to handle whatever is thrown at it.