Nvidia's Icera deal hides bleak outlook for EU wireless

You'll have heard of Nvidia, the graphics and mobile chip maker. Until today, you won't have heard of Icera, which hails from Bristol and makes data chips for mobiles.

So when the former buys the latter, it's Nvidia's story: rightly so, as it fits the fashionable narrative that there's a shakedown in portable tech.

With hot sales but painfully thin margins, every last cent is being squeezed from the supply chain: if you're a sassy chip company that can't offer an all-in-one design with the concomitant economies, you can't compete with the big guys. Icera bulks out Nvidia's ability to add 3G, 4G and other wireless as part of the package and keeps it in the game against monsters like Qualcomm.

That's a nice, simple story. It has the benefit of being true. But it wasn't supposed to end like this.

Back in early 2010, Icera — still not making any money since its 2002 inception — had the PR down to a bullish T. Stan Boland, co-founder and chief executive, told the FT that the company planned a $600 million to $1 billion IPO in 2010, and “We want to build a company worth $10bn to $15bn in the next few years."

Of course he did. At that stage, the company had soaked up some $260 million from European venture capitalists. What Boland told the FT was undoubtedly true as far as it went, but carefully avoided the downside. Without some big deals to back up the plans and break out of dongles, the niche confining Icera at the time, there would be no IPO. With no IPO in sight, an eight year old company in the hole to increasingly impatient corporate sponsors has to find another exit strategy. That means selling out. A year later, that's what happened.

Fortunately, companies with 550-plus patents in wireless technology do still have a decent market value, especially to ambitious mobile chip companies with no wireless to speak of and a bank account bulging from a recent legal knock-out against Intel. The $367 million shelled out by Nvidia seems a fair price, especially since they'll also get some good engineers, some decent designs and a bit more respect in the Far East.

But it's a long way from a billion dollar payday. VCs don't normally aim for 'a fair price' when it comes to collect their winnings; a company that turns $260 million to $367 million over eight years, and that probably at gunpoint from the bankers, is not really what they sign up for. Especially given the state of the market: there simply aren't many Nvidias out there, and next year might have been too late.

Which is bad news for European start-ups looking to use some of the continent's world-class wireless talent, experience and innovatory vim. As the FT noted at the beginning of 2010, there'd been no other semiconductor IPO in the UK since CSR in 2004; as everyone's noticed more recently with Infineon and Nokia's own shotgun weddings to nice Americans with chequebooks, it's a buyer's market.

If not wireless, then where? E-health? Energy management? Both desperately in need of febrile innovation such as has driven mobile technology; both hideously hard for bright kids to break into on their own terms.

One thing's certain: nobody will be talking about billion dollar tech IPOs in the UK for a while yet.