Bold, Hairy, Audacious UK Chip Plays

Bumbling down to the West Country earlier in the week, it seemed to me there has never been a time quite like it. The old companies, Plessey, Ferranti and MEDL stuck to niches: Plessey was a 'creamer' of high end ECL products; Ferranti went for gate arrays, MEDL went for odd things like SOS. To suggest to them that they should do standard product was to engender a look of horror.

There were brief flurries into something more ambitious: Marconi-Elliott Microelectronics and Elliott Automation went into standard logic only to be hastily closed down by Arnold Weinstock in the '71 recession; there was a brief disastrous foray into DRAM and SRAM by STC, and an inspired, but ultimately forlorn, sally into microprocessors and memory by Inmos.

All of these companies had one thing in common. They all got their technology from America.

Icera, Picochip, and XMOS all have home-grown UK technology. And, what's more, they've got generic, broadly technologies which can underpin their transition into major companies.

None of the three companies want the dim and limited fate of making an incremental improvement to a niche product category then getting sold off to a larger company.

They have technologies which, as Icera CEO Stan Boland puts it, allow them to make: "Bold, hairy, audacious plays."

Icera's ability to deliver, in software, superior performance to anyone else at lower power is a technology to take over the wireless industry. Pundits say handset makers won't entrust a handset design to a start-up's baseband chip, but one handset maker already has.

Icera aspires to giant-killing, the giants being Texas Instruments, Qualcomm, Broadcom, Infineon, Mediatek, Freescale and the NXP/ST combo. That's hairy in the extreme.

Peter Claydon, co-founder of picoChip, says picoChip's technology can address "any DSP problem, our aspiration is to move into other DSP market segments". At the moment it's addressing wireless problems, but the scope of so generic a technology is endless.

XMOS has a technology which, according to CEO James Foster, is a 'new industry building block like TTL or FPGA".

TTL built Texas Instruments into a major player; FPGA built Xilinx and Altera into major companies XMOS' SDS could build XMOS into a major semiconductor company. "We're creating a broad-play business", says Foster, "we're building a monster."

Cambridge Silicon Radio has shown that it's possible to build a multi-billion dollar value semiconductor company in the UK, making standard product, using UK-developed technology.

That's something which people used to think was impossible.

Not any more.