What does the IoT mean for the next wave of Silicon Valley innovators? Looking at the previous waves of semiconductor economic development and the doctrine of “creative destruction” holds clues as to how this one develops and who emerges as the new leaders.
Given seven decades of progress, it may seem semiconductor firms on top in mobile would naturally lead the transition to the IoT. The flow of ideas and capital suggests something quite different takes place. Distinct waves of innovation appear, defined early on by shifts in semiconductor technology but later more by advances in speed of adoption and market penetration – socioeconomic effects. These waves are also marked by transitions in leadership.
In a paper titled “Silicon Valley: Global Model or Unique Anomaly” by Doug Henton and Kim Held, the idea of creative destruction as postulated by Joseph Schumpeter is put forth as the pattern for technological change. Schumpeter’s work from 1934 helped define what we know today as the Gartner Hype Cycle, following a technology trigger through inflated expectations and a crash into eventual productivity at a manageable growth rate. That boom-bust cycle is uneven, creating economic upheaval and eventually displacing incumbents who struggle to respond because they are too busy working the old model. They fall prey to the disrupters, free from many constraints.
The technology side of creative destruction may be within the grasp of Silicon Valley planning, but the socioeconomic side is much more unpredictable. Henton and Held cite the seminal book “Technological Revolutions and Financial Capital” by Carlota Perez. Her theory holds that institutions actually inhibit large-scale technology adoption in the early stage. A frenzy develops and runs straight into a multi-faceted wall of incompatible infrastructure, regulation, societal norms, education, and a host of other factors. Resistance on multiple fronts results in a stall wall (a better term than crash, because if the technology has merit it will continue to fly somehow), waiting for conditions across society to adapt and accept the change.
These phases form a single wave. Viewed in hindsight, these waves may be misunderstood, attributed to the list of things taken for granted. Semiconductors have always gotten smaller, faster, and cheaper – at least in the preceding seven decades of collective experience. Products employing them have changed the way we think, work, and behave, and there is no going back assuming we can continue to afford to run these devices and the networks behind them. We can’t remember, or even imagine, a world without them.
What seems like one massively long semiconductor innovation wave is in fact several shorter ones. The composite effect of all the waves reshapes the landscape. Henton and Held use a version of a diagram that many have adapted, showing the timeline of recent technological change powered by semiconductors:
The last wave labeled “social media” is particularly interesting. Many new Silicon Valley firms have risen to fame and fortune on minimizing their own infrastructure and hardware development, focusing on software and leveraging mobile device wielding crowds as content providers. The fabless revolution, with TSMC at the vanguard, enabled the social media wave as smartphones and tablets blazed on the scene.
If the IoT is not the next wave, we are in big trouble. The PC, Internet, and social waves, measured by new investments and acquisitions, are close to spent. Those technologies have reached just about everyone they are going to reach once Asia fills out. The “PC” curtain call, disguised as a 2-in-1, is taking advantage of pent up demand in replacing worn out units and the frustrated tablet users seeking more power – good, but not the growth driver needed.
I prefer to think we are in a forming IoT wave, enabled by sensors, big data, cloud infrastructure, open source software, and the maker movement. VCs and crowdfunders are scouring the earth for the next big thing, but are so far puffing a bunch of little things looking for a niche. The incumbents of the preceding waves find themselves pitted against the same stall wall as the smaller, more obscure firms pursuing the IoT.
Here comes the buzzkill. As Chris Wasden of the University of Utah put it very recently:
— Don Dingee (@L2myowndevices) November 6, 2014
He’s right. Projecting the leading Silicon Valley firms against these innovation waves, finding a firm that dominated throughout two consecutive waves is rare. Where are those guys from the defense and small-scale logic days? TI is the most famous of those names left, bulked up after buying National Semi. Apple is close to sustained wins; they had a hand in shaping the PC market, but didn’t rise to a dominant market share, instead finding their major rebirth in the post-PC era. Intel and Microsoft are also close, leading the PC wave and opening the door for the Internet wave, but they both tragically missed the mobile/social wave by conventional measures.
Real change requires going after high-hanging fruit. Most firms don’t have a ladder that will reach over the wall – and can’t afford to build or buy one. We are beyond the scale of putting consumer electronics in every pair of hands on the planet. The thinking is we are now heading off after the really big problems: smart cities, global warming, clean energy, healthcare, education, those sorts of things. Devices, gateways, and infrastructure in the IoT are going to have to work together.
Paraphrasing Wasden in his full remarks, as of yet we have no example of a firm transitioning leadership from one paradigm to the next. However, there is nothing saying leadership can’t skip generations, catching a wave farther down the line. There is also an important metamorphosis in the supply chain that may break the pattern of discontinuous leadership seen historically.
It is far easier to creatively destroy when you are on the outside looking in than it is to risk messing up a leadership position while trying to catch the next wave. Four companies who missed most of the last wave and had an opportunity to rebuild are off trying to catch the IoT: Intel, IBM, BlackBerry, and Microsoft.
While the headlines have screamed hyperbole like “Intel Quits Mobile” and “Intel Is Not Quitting Mobile”, the focus has been on the futility of mobile investments to date. Less obvious is how Intel has been pursuing their own form of creative destruction, realigning the entire company (not just one division) around the IoT. Daniel Nenni passed me an Intel investor presentation on the IoT containing this slide, a progression showing each manager’s area of responsibility:
Intel is about to cover all three IoT tiers with processor technology – edge, gateway, and infrastructure. Their acquisition of Wind River and support for Android, Linux, and other open source initiatives like the Industrial Internet Consortium and the Open Interconnect Consortium puts them in rather unique territory.
Meanwhile, many people are writing off IBM after they paid GLOBALFOUNDRIES to take over their semiconductor business. Power Architecture has also taken more than a few hits after the Apple scuttling, with Applied Micro adding ARM cores to their stable, and Avago acquiring LSI, leaving Freescale as the last sizable Power.org silicon member standing.
IBM is now all about the cloud. Bluemix is a gigantic platform-as-a-service initiative, not just a bunch of servers with an operating system but instead complete capability for building and running applications. Within their Bluemix is the IBM IoT Foundation, providing a way to ingest device data via MQTT. This gives IBM a big leg up in analytics and context, essential to making any sense out of data from billions of sensors.
It seems odd to say BlackBerry missed the social wave, after singlehandedly creating the mobile email market and building a secure infrastructure. Ridicule away, but they could be one of the great rags-to-riches-to-rags-to-riches stories of the IoT. The acquisition of QNX and their relentless drive toward the connected car has paid off with design wins, most recently snagging Ford as a new customer. Again, we see another example of an entire company realigned for the IoT, this time around Project Ion.
Microsoft is lurking out there as well. Trapped in the difficulty of creative destruction, Microsoft has billions in revenue at stake even as PC and Internet markets bleed out slowly. They can’t completely reorganize away from their heritage. We won’t know the extent of how real Windows 10 is for the IoT until we see the whole strategy sometime in 2015. Support for ARM architecture will be critical, as well as scalability to smaller devices and the ability to compete with open source offerings.
The exception to the pattern may be ARM, having gerrymandered the semiconductor supply chain in their favor and changed the business model. ARM currently owns sensor space thanks to support from multiple MCU vendors, and is riding straight into the IoT wave on the strength of that alone. Their success combined with advances in EDA and IP is drawing competition from everywhere – GPUs, DSPs, FPGAs, and other MCU and SoC cores. The chance they replicate their unfair 95%-ish mobile share in the IoT is infinitesimally small.
Apple and Google have also altered the supply chain, but each is finding the IoT stall wall more difficult than expected. In an environment where everything has to work together, Apple is highly parochial, and Google is the antithesis of privacy. Each will have a measure of success on the IoT, but their long term dominance is unlikely. One model that may work would be to spin off an IoT division; for example, Nest was founded by ex-Apple types, purchased by Google, and may ultimately be freed again.
There is also the thinking that the IoT will be fertile ground for startups. Money will be made, for sure, but staying power will require solving the bigger problems. Most startups will likely flash in, solve a technological problem, and cash out quickly. If a smaller company starts to make big progress on the socioeconomic front, they could cement an IoT leadership position.
Stealing a line from Kurt Sutter: “History doesn’t lie.” While the well-known names are out there, and ARM has a sizable head start, we are likely to see all-new Silicon Valley names appearing and growing to prominence on the IoT. The futures of Intel, IBM, BlackBerry, and Microsoft may well depend on how well their years of planning and massive restructuring pay off with IoT wins. Once the IoT leadership becomes clear as the solutions mature, this will set off a new cycle of creative destruction, where incumbents are forced to reevaluate and adapt – for the next wave.