While Gordon Moore made the prediction in 20th century that the number of chips on transistors would grow double every 2 years, SoftBank Founder Masayoshi Son has now predicted that the number of transistors on a circuit would be more than the number of brain cells in a human by 2018! Computers will outsmart human intellect; means of communications would evolve quick, removing language barrier and even speech barrier! He predicts that telepathy would eventually be used for communication. And these are just few of the many bizarre, yet believable, predictions made so far, leaving players in the semiconductor industry excited, and somewhat confused.
In the past, semiconductor industry has gone through dynamic shifts with a series of disruptive changes, leading to new and higher inflexion points for the industry.
The ‘70s saw the dominance of demand for defense applications giving way to enterprise applications, followed by the consumer devices market explosion in the ‘90s. And this time around, ‘The Internet of Things’ is disrupting the status quo! IoT is evolving fast, from first being a prediction, to being a revolution, and soon to be a way of life.
Source: McKinsey & Company; forecasts derived from ABI Research, expert interviews, Gartner and IDC
But what’s in it for the semiconductor industry? How much of this value-add sales is actually earned by the chipmakers?
As per Gartner’s prediction, the IoT part of semi industry is expected to grow by as much as 36.2% in 2015, while the industry grows at a modest 5.7%. By 2020, this segment is expected to grow by 30% in revenue. This prediction has been made for all existing industries across all end user markets.
It was so far expected that the major sectors being supplied to would be the automotive industry, home automation and medical devices. McKinsey did a further analysis in an attempt to unlock where exactly does the potential of IoT would be, in the new McKinsey Global Institute Report. According to this, the total value added by IoT would be somewhere between $3.9 to $11.1 Trillion. The following is how they predict where most value will be created:
Source: McKinsey; McKinsey Global Institute Analysis
While there are now hints and signals that can be picked up from our industry pundits, there is still no sure shot way of knowing what exactly is going to work when and how would it eventually pan out. According to Joe van Beurden, CEO of CSR (a leading fabless semi conductor company) it is still tough, as of now, to put a finger on one or more industries that will actually be contributing to the exponential growth in the semiconductor industry. The major reason for this uncertainty is that disruption is not necessarily being caused in the semi conductor industry, as far as IoT is concerned, and instead is following disruptive applications of the “things” in those particular industries.
Successful application of IoT in every industry would require a method to the avalanche of data collected and stored. More so, in such a manner that it can be crafted into data that’s usable for individuals, organizations or devices, to bring that value addition in the end user’s life.
Now that’s not an easy job, considering the uniformity required at each stage of the technology adopted. And lastly, with collection of personal data comes the issue of privacy. How an industry copes with that will also determine the ease of execution of IoT technology and applications.
While challenges are casting a shadow on IoTs in 2015, sentiments still are positive. Demand for first generation of products is definitely there, and backing this statement is the estimation of industry pundits – almost 3 million Apple Watches are expected to have been sold in just the first 3 months of its launch! That’s a statement in itself. And the semiconductor industry seems to be catching up as well in terms of technology advances suitable for these evident opportunities. For instance, more advanced processors (like ARM’s Cortex M) are being manufactured, which are almost 10 times more efficient and can be used in smart watches. Besides wearables, Gartner predicts a hike in demand from Automotive Industry, LED Lighting and Home Consumer Products, for mostly processing, sensing and communications devices.
IoT is the future without doubt, and has not just billions, possibly trillions, of additional revenue globally. Who all benefit in the semiconductor industry from this will depend on who has taken the most calculated risk, understood the end user’s future needs in time, and adapted to the changes in various segments of the market, most profitably. Easier said than done, for sure, but the prize at the end would be well worth the effort!
It’s not just the Internet of Things that’s keeping leaders on their toes, but also the increase in competition brought about by emergence of new companies, in this otherwise oligopolistic market.
Similarly, Intel, the alpha of semis, has been failing to make its mark in the mobile chip business. It’s not like they haven’t tried, but have been making poor decisions along the way. They refused to make chips for iPhone back in 2005, because the top executives did not expect iPhones to gain popularity. As a result, Apple, which at that time was only a year old in the mobile market, stepped into development of chips themselves. They created a design team, poached employees from leading chip makers (like Texas Instruments), and eventually not only matched itself to the leading companies in this industry, but even went ahead of them with innovation in mobile chips. And what’s more, if rumors are to be believed, Apple would soon have its own in-house design team to produce x86 chips for Macs (currently supplied by Intel) and 4G baseband chips for iPhones (currently supplied by Qualcomm).
Apple successfully replaced AMD and Nvidia, their original GPU sources, by getting its in-house team to custom design a powerful GPU based on Imagination Technologies PowerVR mobile graphics processor IP, just as it had done decades earlier by going for ARM’s mobile processor IP for its iPhones.
The long-standing GPU duopoly lost an important contract and won a strong competitor! The impact has prompted Nvidia to leave the mobile market and shift to the automobile sector in a big way.
Needless to say, the traditional leaders of the industry do see upcoming companies as a threat, probably the reason why there have been a bunch of acquisitions and mergers by the bigger companies.
NXP semiconductors announced in March 2015 that it would be acquiring ‘Freescale Semiconductor,’ its competitor, which led the stock prices of NXP to rise by more than 17% in March itself. The automotive semi conductor industry has especially been abuzz with mergers and acquisitions. In June 2014 ON Semiconductor acquired Aptina Imaging corp for $400 million, and Infineon Technologies acquired International Rectifier for $3 billion. In October 2014, Qualcomm bought a fabless semiconductor company called CSR PLC for $2.5 billion.
Additionally, according to a report by CNBC, “analysts see companies that make chips for connected devices as top contenders for deal activity.” That brings us back to IoT. According to the same report, it mentions that the level of M&A is expected to catch pace, with deals’ worth ranging from $600 million to $3 billion. Clearly what we’ve seen so far is just the tip of the iceberg!
While objective of acquisitions is to reduce competition, such an activity should ideally add more value, like sharing of technology, customer base, cost efficiencies through scale, or reducing cost by backward or forward integration in the value chain.
The challenge for bigger companies is to identify which such start ups globally should be tapped into. And it’s not that easy. Main reason being that bigger companies like Intel, Qualcomm, Samsung have products and technologies across industries, while most of the smaller companies produce more focused products and technologies.
Which ones to go for, and which ones should be resisted from buying, is a tough call for the top executives, and requires experience, understanding of all the businesses in the value chain completely, and a good foresight into the future needs. And adding to that, the volatility and changing demands inherent in the semiconductor industry, does not make their jobs any easier.
Changing demands, technology innovations, and more importantly the need to improve profitability has led disruptions across the value chain. For instance, the semiconductor companies have decoupled most parts of the manufacturing process, and outsourced it to foundries, to reduce capital-intensive investments. This includes even Japanese companies like Toshiba and Fujitsu, who are known to keep everything in-house.
The foundry business has thus strengthened, growing consistently by 5 percentage points each year as per McKinsey’s report. It is proved to be a successful model, where everyone’s a winner.
Its growth has allowed key players to focus more on core capabilities and evolve their current technologies to match pace with the changing needs of industries, and not to forget, the needs of IoT, which cannot be fulfilled without persistent focus in improving the core technology. The foundries too are not standing still. The biggies like TSMC are adding design and packaging capabilities so that they can provide more value-added, higher-margin, services to their OEMs. This also enables new entrants, especially the non-conventional ones, to launch new products with far less capital investment.
Packaging, which used to be an after-thought with most chip designers, has suddenly shot into prominence as the cost of adhering to Moore’s law has become more and more prohibitive. The emergence of consumer facing products too has added to the pressure of innovating on packaging to leverage the real estate better and also manage the heat/power issues. The results are already showing in the recent products in consumer hands.
Most recently, Google is revolutionizing the hardware segment of mobile significantly, with the ambitious Project Ara, and with that giving the semis one of the biggest challenges (or opportunity?). Ara will be empowering end users with the option of creating a smart phone that’s exactly how they would want. So, this way, an end user can have different memory spaces (not just the option of 16GB or 32 GB) depending on their need, the exact kind of camera, it could have sensors for applications relevant for the user. It also lets you work out aesthetics of the phone, as you’d want. So now, electronic companies no longer need to guess what’s best to be produced and instead, be part of an ecosystem wherein the customers say what they want and the same is delivered with ease.
Players in the semiconductor industry should definitely see Ara as an opportunity to innovate their product and value chains. Not only the big players, but also SMEs and start-ups should be leveraging this opportunity.
Key takeaway from this for CEOs is that there’s always scope for optimization in the value chain, by either re-allocation of investments, finding the perfect collaboration, or decoupling of processes. The challenge for leaders is to know where to look exactly and be open to change. Or more precisely to hunt for change that can help their business evolve and give them an edge over competitors, from within as well as outside the industry, and an upper hand in the technology race.
If you are in the semi conductor industry, you’re probably aware of Gordon Moore’s prediction in 1965, later coined as Moore’s Law, that every 2 years (later changed to 18 months), the number of transistors on a given integrated circuit would double, gradually incurring lower costs from economies of scale. Moore’s Law has given direction and focus to research teams in the industry, and has proven to be true for the past 5 decades. However, it is being debated now if its time to let go of this law, probably.
There are 2 important reasons for this. Firstly, Moore’s Law works more as a supply side strategy, where it works at only optimizing on costs. How should innovations be made based on changing market needs is not addressed by the law.
Secondly, Moore’s Law is relevant for microprocessors and memory. But there are some essential diversified technologies, that provide disrupting value addition to devices, to which Moore’s Law does not apply to, called as More than Moore technologies. It includes radio frequency (RF) devices, power management subsystems, passive components, biochips, and sensors, to name a few. The theory of ‘More than Moore’ is more relevant for IoTs as well, which does not hold ground when seen only from the point of view of Moore’s Law.
This means that decision makers would need to ignore the tried and tested law while taking strategic decisions for their organization. Clearly the leadership needs to be bold enough to be able to let go of what’s worked in the past to be able to see the future in fresh perspective.
The uncertainty in this industry could not be more evident. The big players can already feel the heat of a more competitive market. And with IoTs, the pie is expected to grow much bigger. Who will get what portion of the pie is yet to unfold.
To avoid being left out, businesses need to re-establish their hold in the market given the sweeping changes occurring across all major industry markets. Companies will need to focus on doing better, what they do best, and make a dent in the target industry before a competitor gets there. For a lot of companies, a complete re-wiring of the business will be needed to tap the potential of their market.
It is going to be mighty interesting to watch who the winners are going to be, who the has-beens, and how this entire transformation within the semiconductor industry plays out over the next few years. But one thing is certain. Who wins or loses, is going to be determined by who is at the helm of these companies. Conventional leaders won’t do. The industry is going to need a whole different set of leaders in the years to come – Leaders who are passionate, entrepreneurial, and think out-of-the box. Leaders who have a vision for where the industry is headed, where their company needs to go, and know how to lead from the front. Leaders who know how to take learnings from one industry and market, and apply them in another one. Most importantly, Leaders who know how to take the strategic risks, challenge the status quo, and succeed. In short, there is going to be a big premium for Transformational Leaders!
Get instant blog updates by email
Have a question about AnuPartha?
We've got it all covered.