VC investment fell nearly across the board in 2012, but the electronics industry remains especially neglected.
Some of the lowered funding is due to lack of interest or the
perceived long-return windows or frustrations about the global
macroeconomic environment. But some of it undoubtedly is due to the
fact that semiconductor start ups are leaner than they were 15 years
ago, fewer engineers, better tools delivering more productivity,
more robust global supply chain to leverage. In other words,
semiconductor startups are less expensive to invest in.
It's time to move past denial. The good news is engineers are
getting amazingly creative in bringing their creative ideas to life.
Exhibit one is our friend Andreas Olofsson and his very small team
at Adapteva, which recently crowd-sourced
$900,000 on Kickstarter to fund a new mask set for its
Epiphany processor to build the next generation of their
signal-processing device. Even before that, he'd already done many-core
processors at 65- and 28-nm for just $2 million.
The other trend I cling to is that all the amazing innovation of the
past decades in hardware and software is not only putting
accessible, affordable hardware and software tools in the creators'
hands, it's enabling more and more affordable applications for the
consumer market. Remember, this is a relatively new opportunity for
an industry that has long been a slave to behemoth segments like
mil/aero, computing and telecommunications.
When a kid can conceive of aftermarket driver-assist system that
uses low cost, stick-on sensors communicating via WiFI with your
smart phone or tablet, it's a whole new ballgame (I will
find this guy). And it's a ballgame that allows more engineers and
entrepreneurs to play without the VC stranglehold. Greater risk
perhaps, but more freedom, control and potential reward.
Valley Nation: You want how much?!
to make a processor for (way) less than $100 million
Valley Nation: Forbes innovation list disses electronics
Kickstarter created a community for Adapteva
Kickstarts a $100 supercomputer