Something’s going on in the UK hardware scene, though it’s yet to be captured by any aggregate statistics I know of – as always, where’s a journalist when you really need one?
What I mean is: for a while, it has felt as though my company has been one of the very few ‘proper hardware’ UK tech startups (medical / energy aside) out there, but now I’m hearing about an upswell in UK electronics design work that began mid-2010, and that began to go seriously big over the last month or so.
Sure, a lot of that is bound to be part of the semiconductor sector’s overall bounce back: and the move by chip suppliers to put out small pitch (i.e 0.5mm) mobile-class SoC devices (which you (a) laser drill or (b) give up on and go home) has meant that low-end pooling has ceased to be an option for a lot of new boards. All the same, it does indeed seem that UK companies are now actually building stuff (shock, horror, etc).
Perhaps all this is a leading indicator of a flurry of hardware startups that will reveal themselves in early 2011, or just budgets shifting back from present-tense survival to future-tense R&D: I guess we’ll see soon enough. Yet even though I’ve said for a long time that hardware is the new software (i.e. that there’s much more of a business case to be made for investing in defensible rapid-prototyped hardware plays than in indefensible Internet software plays), that remains my personal contrarian position, and still a couple of years from becoming accepted as mainstream investment thought. Though… feel free to catch up with me when you’re good and ready, guys.
As an aside, there’s something seriously wrong with 0.5mm pitch between pins – though it’s great for mobile devices (which need to be ultra-compact and ultra-mass-market), it’s a lousy design choice for everyone else. Even though the same chip with 0.8mm pitch would occupy ~2.5x the space, the space saving comes at the cost of layout agony, of fabricating multiple layers, of laser drilling, of bonding layers together, etc. The production issue is that much as the whole point of manufacturing is to design something and then build a million of them (don’t get me started on ‘niche’), applying small-batch lean manufacturing mucks up the whole economics model assumed by 0.5mm. But I’m clearly getting too technical here, so I’d better stop.