Why Are There Fewer Chip Start-Up Successes?

| No Comments | No TrackBacks

Rahul Sud, former Inmos SRAM designer and founding CEO of Lattice Semiconductor, asked one of his fiendishly unanswerable questions at the panel session at Future Horizons' System and SOC Conference in Prague yesterday.

“If you look at the semiconductor start-ups for 1983 to 1993 there were many successful companies, but if you look at the 1993-2007 period there have not been many successful companies”, said Sud, “was it that the VCs of the 1980s were really competent, or did all the VCs suddenly lose their marbles in the 1990s, or did something in the industry change?”

“Yes, something changed,” responded Stan Boland, CEO of Icera Semiconductor, “the bar got raised.”

I suggested that although there were some radical new departures in the industry in the 1980s like CMOS programmable logic, and the fabless business model, there were no such radical new departures in the 1990s. Was this because no one had any Big Ideas in the 1990s or because, although there were some Big Ideas, VCs wouldn’t fund them?

The response of James Foster, CEO of XMOS Semiconductor, basically was that the VCs wouldn’t fund big risks any more. “Because costs are going up, the risks venture capitalists will take are going down,” said Foster.

While venture capitalist Martin Gibson of Atlas Ventures thought: “Almost all the companies we back are tightly focused.”

An exception to that is, of course, the previous speaker from XMOS who’s software defined silicon ICs are aimed at everytrhing consumer. You can’t get much more unfocussed than that.

Sud then inserted another of his stilettos: “The FSA (Fabless Semiconductor Association) reports that there are about 140 semiconductor start-ups every year each taking between $30m and $120m. Over the last 15 years that’s 140 X $70m X 15 which is the amount of money that’s gone into the semiconductor start-ups. Where are the profits?”

“All the profits are concentrated in relatively few companies and the rest are moderately profitable or fail”, responded Gibson.

Of course whether or not the semiconductor industry makes profits, or has ever made profits, is one of those hoary imponderables.

It’s a bit like asking: Which came first, the chicken or the egg? Or: What does the crocodile have for dinner?

Best left unanswered, or better still, unasked. Rather poor taste.

No TrackBacks

TrackBack URL: http://www.electronicsweekly.com/cgi-bin/mt/mt-tb.cgi/13034

Leave a comment

Get the eNewsletter

Sign up for the weekly Mannerisms eNewsletter. Get the blog highlights straight to your email inbox, Tuesday morning, no fuss. Just tick the option for Semiconductor commentary.

Archives

Get Mannerisms via RSS

OpenID accepted here Learn more about OpenID

Sponsored by Mouser

Sponsored by Mouser Mannerisms is brought to you in association with Mouser.

Recent Comments

Advertisement


Sponsored by Mouser

Sponsored by Mouser Mannerisms is brought to you in association with Mouser.