Offerings of novel computational architectures, in the form of machine learning and open hardware, mark the 18th rendition of EE Times' Silicon 60.
The geographical trends in the Silicon 60 that we have identified in previous years have been maintained in 2016 and 2017. California and China continue to be hotbeds of startup creation and getting hotter. Some other regions struggle to breakthrough from startup creation to the point of producing major market-making and wealth-creating companies.
In this year's list – v18.0 – The United States provides 31 of the companies, a decrease of one from the previous year. The state of California provides the headquarters for 25 of the 60 startups. This is up from 21 in 2016 and a low point of 15 companies in 2014. With one Canadian representative on the list that gives North America a count of 32.
The rise and rise of California is based on a number of factors. As the location of the original Silicon Valley it has a first mover and critical mass advantage. It is also home to Sand Hill Road and the nominal center of venture capital. Silicon Valley or the US at large is often a location of choice for companies' second offices. And often that second office becomes the headquarters even when the bulk of engineering employment is elsewhere in the world. The return of venture capital to the hardware sector could also have something to do with a spate of startups back in the Valley and its outer fringes.
China, is the second-placed nation state with five companies on this year's list, up from four. Elsewhere in Asia, India has maintained its count at two and Japan, not well known for startups, now also has two.
Traditionally Japan has had a culture of extreme stability in professional life with employees frequently spending all their working lives in one of the major vertically integrated corporations. The US style of multiple short-term, project-based stints of employment for specialist companies in a disaggregated supply chain was…foreign. However, with globalization and international competition such cultural differences are diminishing.
And, Japan has struggled with the profitability of its major electronics corporations. The long-running unfortunate saga of Toshiba in the last couple of years has been the most recent case in point (see Politics Muddy Toshiba Bidding ).
We therefore predict that gradually Japan, which has excellent academic research into electronics and semiconductors, will gradually produce more startups. How long it will take for one or more of them to break through to world significance is another matter. Asia has a total of nine companies in this year's Silicon 60.
Meanwhile in Europe the unedifying process of Brexit continues to tear at the continent and has serious short-term implications in terms of limiting economic growth for both the United Kingdom and the European Union.
But in startup-land hope springs eternal. In Europe funding levels tend to be less than in North America where there is more money available and the attitude is more likely to be "go big or go home." Europe is good at producing interesting technologies and getting companies started. As we have commented before it also has a number of state-supported research institutes that have as a remit the need to push researchers out of the nest and form startups.
This is now making for lots of startups in Europe but with the risk of too little investment in technologies that are looking for a problem to solve, rather than the US model of larger investment in solutions to a problem with clear market traction.
Nonetheless Europe is home to 14 companies on the Silicon 60; three each from France and the UK, and two each from Finland and Switzerland. The is an increase of one from 2016. Four companies based in a traditional hot-bed of innovation, Israel, and one company from South Africa completes the geographical spread.
With this year's influx of young companies the average age of the startups has come down from four to three. Exactly half the Silicon 60 was founded in 2014 or later and half was founded in 2013 or earlier.
We spoke last year of the electronics and semiconductor industries taking a collective breath before diving into the next big thing after the smartphone. And we indicated it might be found somewhere between autonomous driving and the machine learning. As multiple startups take the plunge — alongside major players or abetted by them — it would seem that our guess was not too wide of the mark.
The need to reduce the power consumption of complex computation by bringing capabilities on to the leaf-node equipment is being met by getting them to mimic biological systems. However, even while this revolution plays out one wonders who will pioneer the novel materials and active devices that could augment or replace silicon as a platform for such biological style electronics, or spintronics.
It may be that these developments will, by necessity, remain in the control of the larger companies. However, these that have become large and successful in the era of CMOS are not always the most highly motivated to usher in fundamental change.
Therefore, in these areas and many others, startup companies will continue to serve a pioneering role — and the Silicon 60 will be here to observe and celebrate them.
— Peter Clarke writes about the semiconductor industry and was formerly an editor at EE Times.