I wanted to thank the SSIA to invite me to share with you some of my thoughts and views of the semiconductor industry.
To start, let me recall that I am a veteran of this industry. I started working in March 1963, just after I received my master in Electronics at the Politecnico of Turin, as a salesman of the distributor of Motorola Semiconductors in Italy, and I retired in March 2005 with my last employment position as president and CEO of STMicroelectronics. All during those 42 years, I basically worked only for 2 companies: Motorola, initially as an employee of the distributor and then directly as their employee until my last position in 1980 as General Manager of the International Semiconductor Division based in Phoenix Arizona. The other 25 years were spent at STMicroelectronics initially as CEO of SGS, one of the two parent companies of ST, from 1980 to 1987 , and then as President and CEO of STMicroelectronics until my retirement in 2005.
In all those years I worked and lived in several countries including Italy, Switzerland, the US, Italy again, France, and finally Switzerland and I travelled, as it is normal, extensively throughout the world including in Asia and particularly in Singapore, that played a central role in our globalization strategy. And even after I retired in 2005, at the age of 69, I remained associated to the industry both as Honorary Chairman of STMicroelectronics (a title that implies moral association but without any business responsibilities, except for the first year when I acted as a consultant ) and because I served in the board of Chartered Semiconductor, a company based in Singapore until it was sold at the end of 2009 to Global Foundry. Therefore, I have practically lived through nearly the entire life of this young industry.
As you know, all started with the revolutionary invention of the transistor in the Bell Labs in 1948. The 50s saw the first pioneer starts of industrial activities (for example the companies I worked for were born in that decade, Thomson Semiconductors in 1956, SGS in ’57, and Motorola Semiconductors in 1958), and from the 60s on, the industry was accelerating its growth and pervasiveness of every aspect of human activity and life.
During all those years, the industry has experienced periodic corrections that have been and will continue to be the way of life for the next several decades until the industry will continue to be dramatically innovative and far from maturity.
Those corrections are caused by years of rapid expansion associated with excessive capital investments and inventory build-up, followed by sharp decline induced by those imbalances or by economic slowdowns or both. And always are impacted by the fast rate of innovation.
Those corrections have normally occurred twice in each of the past five decades with large swings from the growth of the peak year to the decline of the bottom year. The most severe swing in the industry history was in 2000/2001 moving from a growth of 37% in the year 2000 to a decline of 32% in 2001 with a swing of 69%. In comparison, the last recent swing of 2008/2009 with 41% was relatively moderate.
Now, we have entered in one of those corrections. So what? It is business as usual in our industry.
What is the swing likely to be this time?
In early October, I attended the yearly forum of Future Horizons which is an outstanding "Global Semiconductor Industry Analyst" based in London, UK. This year the forum was held in Seville, Spain. Because I know them very well and they have been historically among the most accurate in their forecasts, I will report their opinion.
First of all, they stated that this correction is not induced by severe imbalances in the industry, like excess capital investments (as it was the case in the 2000/2001 severe correction) but rather by a world economic slowdown that was not anticipated. In April of this year, the IMF forecasted a world GDP growth of 4.5 % both for 2011 and 2012, while in September it had been revised downward to 4.0% for both years - the slowdown essentially due to Europe and US: significant but not dramatic. This assumes that any major financial crisis in Europe or US and consequent recession can be avoided.
Under this economic scenario, Future Horizons expects that when all the numbers are in, Q3 of this year will be basically flat with Q2, Q4 will be moderately lower versus Q3, bringing the entire growth of the year 2011 to +1% versus the +32% of last year’s growth and therefore a moderate swing of 31%.
Next year, after a weak start in the first half, they expect (in the above economic scenario) a strong rebound because not only the capital investments in the industry in the last several years have not been excessive, but Future Horizons consider those investments as being below the long term trend line. This rebound will bring the total growth of 2012 to 8% followed by a strong 22% growth in 2013 with the semiconductor industry reaching the level of 400 billion US$.
As far as I am concerned, I agree with the above scenario, but since I believe that the economic slowdown will be more severe than what actually forecasted (I also believe that a major crisis will be avoided in Europe and USA but both economies are burdened by huge debt issues as is Japan and also the BRICS countries are having some economic challenges of their own) and therefore this year’s swing in our industry may end up being a little bit bigger and next year’s growth smaller. However, nothing fundamentally changes.
What I recommend to my younger colleagues, is to act, as I think should be done in every market correction:
- Don’t panic. Cut rationally expenditures – not linearly. Control inventory and capital investments but don’t sacrifice any critical capital investment for short term benefits. The cost of having some extra capacity unused if the recovery arrives later than expected, is much much less important than the cost of not having capacity in place to serve your customers when the recovery arrives therefore hurting them and damaging both your market share and your credibility.
- Push R&D. During the correction phase, any major R&D program must be accelerated, not delayed: the exit from the correction is faster and more profitable via the introduction of new technologies and new products.
- Go shopping. Any correction phase presents also good opportunities because the strong corporations can go and acquire assets from weaker companies at convenient prices or acquire competitors altogether.
- Communicate. During a correction, you and your senior team must step up the communication efforts with your people and your customers in one side to keep your people well informed and involved to keep their morale up, and to keep your customers aware of your actions and plans as well as to get from them a better feeling of where them and therefore the industry is going.
Where is the industry is going?
In each of the past five decades, a new major application driver has emerged to add to the growth of the established drivers.
In the 60s, the drivers were the big computer mainframes and the telecommunications networks, addressing the productivity of the infrastructures.
In the 70s, the new driver was the consumer electronics addressing the entertainment of individuals.
In the 80s, the new additional driver was the personal computer addressing the productivity of the corporations.
In the 90s, the new additional driver was the wireless communications addressing the productivity of both corporations and individuals as well as the personal needs of easy access to communication.
In the 2000s, the decade just concluded, the major new drivers have been the web and the convergence of consumer and computer and communications addressing the productivity of corporations and individuals as well as expanding enormously the ability of individuals to communicate between themselves and to have huge access to information and entertainment.
In the current decade, the new major drivers will be health care, energy management, safety and security, addressing a broad variety of societal needs.
And as far as I am concerned, I believe that in the next decade, the 2020s, the major new drivers will be artificial intelligence and robotics addressing the broadest needs of society at large.
So, the rate of innovation will continue aggressively, offering new opportunities to the semiconductor industry players.
From a manufacturing point of view, we have moved from the 6 inch wafer size of the 70s to the 12 inch size of today and by the end of this decade the 18 inch will enter mass production. At the same time, the technology node has moved from the micrometer of the 80s to the 32\28 nanometer of today and it will be well below 20 nanometers at the end of this decade ( Intel is reputed to be ready with 14 nanometer by year 2015). At the same time, a major diversification in the chip technology has emerged in the 90s, and is in full evidence today and will become absolutely fundamental during this decade and afterwards.
In one side, we have the "More Moore" field driven by the constant race to finer geometries and in the other hand, the "More than Moore" field characterized by diverse technologies like analog, sensors, intelligent power, embedded micros and RF.
The two fields have already very strong differentiating characteristics. The "More Moore" field, is and will continue to move towards finer geometries, larger wafer sizes, complex system oriented CMOS platforms that require hundreds or thousands of engineers, and therefore presenting a very high and growing financial entry barrier that will force and has already been forcing the consolidation of this part of the industry.
The memory field presents the same technological characteristics and is included in this category.
In this "More Moore" camp, there will be very few players in each major application with the leader reaping up the major profits, the next two followers participating profitably, and the others struggling to survive and possibly disappearing. To be a player in this field, you need to have a big dimension of scale, in order to support the huge investments in R&D for the development of the platforms and the capital investment for the giga dollar fabs or to command good prices from the fewer and fewer foundries that will be available in this field.
For example, today there are only three independent foundries with 12 inch capability with Samsung participating in the foundry business as a non core activity.
Moreover, the life cycle of any new major platform is becoming shorter and the costs of missing one generation are huge.
In the field of "More than Moore", the game is totally different. The needed technology is usually some or several nodes behind, the needed fabs are much more mature and less expensive, and the available foundries are numerous. You don’t need armies of engineers working at very complex platforms but very brilliant designers in relatively small teams working at ingenuous new devices addressing a growing number of applications and the designers must be intimately associated with the application. The consolidation process in this field is much slower because there is plenty of room for new entry due to the low cost barrier to enter the field and the continuous emergence of new applications in many new areas addressing home automation, building automation, energy savings, energy management and control, sensing and measuring, machine to machine communication and personal needs in the area of health care safety and security. In this field, the opportunities are huge not only for the big boys but also for the small medium sized creative companies or new entries with new innovative ideas.
Corporations in both campuses, with the different characteristics I listed above, will continue to grow and prosper if they have the competencies to address their field and will contribute to the solution of many problems and challenges of human society.
However, as always, the key ingredients for success are associated with the human factor. The 20th of august of this year I was awarded from the IEEE the Robert Noyce Medal to the career for the achievements of STMicroelectronics under my tenure. After the reading of the motivation by the president of IEEE and presentation of a profile of the award winners by a prominent industry leader, the nominee was given only one minute for his thanking address. In my minute I thanked all my 50,000 colleagues worldwide that have made STMicroelectronics great and I synthesized the elements of our success in three points:
- Innovation, pervasive innovation, both incremental and disruptive as a way of life of the Corporation and the innovation being driven by strategic alliances with key customers.
- Our globalisation drive with Asia playing a major role and Singpapore at the center of our Asian strategy. Our globalization concept was not only driven by cost considerations but especially to make us closer to our customers in the world.
- Our Corporate Culture which can be synthesized in three concepts
- Integrity at all levels: people integrity, business integrity, corporate integrity
- The pursuit of Excellence by every individual and by every business unit of the Company as a way of life including our TQEM journey (this gave us many recognitions and awards like the the European Quality Award in 1997 , the Malcolm Baldridge Award in the year 1999 and the Singapore Quality Award in the same year; and to be members of the DJSI, Dow Jones Sustainability Index, since the beginning of this index)
- Social responsibility, including environmental responsibility in which we were a pioneer and we have always been an industry leader.
For me, the roles of the leader, not only the CEO but every manager of a business unit, are five:
- Create the vision
- Build the team
- Install the culture
- Design the roadmap
- Drive the execution
Going forward, the semiconductor industry will continue to be characterized by five connotations:
- Innovation: both incremental innovation and disruptive innovation. And also not only innovation in the technology, products and manufacturing processes, but also in the operating processes of the company, ie. in the way the company organizes itself and copes with the market.
- Globalization: it was one of the first industries to go global
- The geographical importance of Asia as the motor of growth in the next 20 years (a major element of the success of ST was to understand very early this factor - we were the first company to start a fab in Singapore in 1984 ten years before Texas Instruments, we were the first to establish here a design center, and we were among the very first to establish design and manufacturing activities in China and very large design activities in India).
- The time factor. Speed is essential. The innovation speed is breathtaking. There are very strong rates of innovation and shrinking lifecycle of the products. Companies must be obsessed with time to market and must organize decentralizing as much as possible the decision making responsibility so to make the decision making process and execution fast.
- Social responsibility. As for any other business sector, social responsibility has been important but now it has become a must under the pressure of employees, customers, citizens and governments. Companies that don’t have a very strong attitude in social responsibility will not be able to compete.