10 January 2012
Is the end of electronics as we know it?
Picochip's acquisition by Mindspeed crosses another company from the list of UK semiconductor businesses. In the comms arena alone, Innovision, Icera and Picochip have been acquired by US companies in the last year or so.
In some ways, this can be regarded as the natural flow of things; companies are born, grown and get bought, replaced by aggressive start ups exploiting new market niches.
But that's a model that worked some years ago. Today, venture capitalists are reluctant to invest in electronics companies, even if they have funds to work with, because there is no apparent return.
Have we, then, seen the end of the electronics industry as we knew it – the rush to design and manufacture at the 'bleeding edge'? Probably. High volume consumer asics will continue to be made on 'bleeding edge' processes, as will the latest fpgas, but unless volumes can be guaranteed to be huge, it will be crazy to consider 20nm and beyond. Industry thinking is turning to embrace a raft of legacy processes which may well be quite suitable, if not quite so macho.
The general agreement is the 'going rate' to get a start up to silicon on a leading edge process is well in excess of $100million. While that kind of cash is available from investors, it won't go into an electronics company because there returns aren't there. Might a return to legacy processes, combined with such approaches as systems in package, change the economics such that investors look upon the electronics industry more benevolently in the future?