Major Industry Trends
If there is one sector that is changing the nature of the world we live in faster than any other, it is electronics. From mobile phones to the internet, from telecommunications to satellite TV, electronics are ubiquitous, and advancing in leaps and bounds, with more power being crammed into less space year-on-year. It is fashionable at this point to quote Moore’s Law. Geoffrey Moore was the cofounder of the PC chip company Intel, and the man famous for predicting that the number of transistors on the same-sized piece of silicon would double every two years—without bothering even to imply an end date for this process.
A few years ago, it was thought that Moore’s Law was running out of steam, as circuits were (a) becoming small enough for quantum effects to introduce instability in current flow, and (b) becoming crowded enough for the heat generated from operating the chip to start to be a real problem. However, advances in silicon substrate technology (the introduction of metal oxide gates, for example) opened up the door again, and Moore’s Law still holds good. Instead of Intel merely producing one central processing unit (CPU, the calculating “heart” of a microcomputer) on a chip, it produced first two CPUs per chip, then four, then eight, and we are very close to 16-CPU chips, with higher multiples possible and planned beyond this.
Simultaneously, the software industry is on a steep learning curve as it rewrites its applications to take full advantage of the vast amounts of processing power becoming available, whether at the server, on people’s desks, or on a plethora of handheld devices, from palmtop computers to mobile phones and tablet PCs. These were first popularized by Apple’s iPad, but by 2011 virtually every hardware manufacturer seemed to be producing a tablet PC.
As the amount of cheap computer power available to engineers has increased, the power of electronics to transform the world has moved forward, with no part of industry now untouched. The ability to simulate real physics inside “the box” has allowed car makers to stress-test both virtual parts and the whole design, long before metal goes anywhere near being machined. In advanced medicine, biosciences companies model molecules and processes to predict drug interactions on target proteins or cell constituents, before any real-world work is done. In the oil-and-gas sector, vast data sets from seismic and advanced scanning of reservoirs are turned into visual, three-dimensional models that geologists can “walk through,” to examine reservoirs “from the inside” before any well is drilled.
The ability to create and explore real physics through virtual models increases dramatically with each new breakthrough in processing power. Advances in electronics truly have the power to change the rules for whole industries.
The electronics manufacturing sector is generally separated from the electrical manufacturing sector by a technical distinction. Under this, the term “electronics” refers to the flow of charge through non-metal conductors, such as silicon in semiconductor implementations, and “electrical” refers to the flow of charge through metal conductors. Electrical is all about wires, and electronics is all about semiconductors, broadly speaking. The latter leads to printed circuit boards (PCBs) and memory chips, while the former leads to white goods and power stations (with the proviso that almost all electrical goods these days have some PCB control circuitry somewhere).
From an industry perspective, electronics has spawned a vast range of specialist industries, from the IT industry (dominated by the likes of IBM and its competitors at the mainframe end, and Intel and AMD at the PC end) to the mobile-phone market. It includes TV-set manufacturers, video-game consoles, and, across an array of industry sectors, a vast army of specialist control-systems manufacturers, not to mention the aviation and automobile sectors, both of which could not exist in their present forms without massive input from the electronics sector. Then there is the medical devices market, and so on.
Asia, today, is regarded as the powerhouse of PCB and memory-chip production, yet it all began through a combination of some breakthrough research in Japan, and US companies outsourcing first PCB assembly, then PCB fabrication, to Asia to take advantage of cheap labor rates. Today, Asia’s dominance in motherboard and memory-chip manufacturing has reached the point where a US air force colonel, Charles Howe, was prompted to write a strategic study looking at the potential impact on US national security of having so much of the electronics industry outside the United States.
The point is not without irony, as the semiconductor industry began in the United States with innovations such as Texas Instruments’ invention of the integrated circuit in 1958, and Intel’s production of the first 8-bit microprocessor, the 8008, in 1972. Howe’s study gives a very clear account of the integrated circuit (IC) design process. The basic design process requires electronic design automation (EDA) software in the hands of experts. Fabrication involves “etching” or imprinting the designs onto silicon wafers.
Each new generation of chip tends to require either a totally retooled fabrication plant (“fab”), or a new plant built from scratch. Each plant costs around US$3 billion, which means that each new generation of chip represents a huge bet by the manufacturer that it can sell vast numbers of its chips to the various global markets.
Very few companies in the world can bet on that scale and get it wrong twice, so semiconductor chip manufacturing is a game with very high entry costs, and is played for very high stakes. Asian chip companies generally play a safer game, and focus on producing not CPUs, but peripheral components, such as motherboards and memory chips.
The entire semiconductor industry is tremendously vulnerable to downturns in the economy, because in all markets, from PCs to mobile phones, the sale of new products is predicated on global growth. There is no doubt that the semiconductor industry suffered heavily from the global downturn of 2008 as both businesses and consumers cut their discretionary spending. For businesses this meant cutting back on IT projects and for consumers it frequently meant economies such as not replacing last year’s mobile phone with this year’s latest model. The impact on sales was marked, with the sector recording its first year-on-year drop in sales at the end of 2008 since the dot.com crash of 2001.
In June 2011, the World Semiconductor Trade Statistics (WSTS) organization released its forecast for semiconductor sales going through to 2013. Based on sales for the first half of 2011 the WSTS forecast growth of 5.4% for the sector in 2011, with total sales worth £314.4 billion. This represents a 1% increase on the WSTS forecast for 2011 issued in November 2010. The figures are generated by the WSTS from an extensive group of global semiconductor companies. According to the US Semiconductor Industry Association, the WSTS forecast now looks exactly right, with the industry expecting moderate year-on-year growth through to 2013. This is being driven by steady global demand for high-end electronics, boosted by increasing demand from emerging economies. The WSTS is anticipating growth of 7.6% for 2012, with growth then moderating back to 5.4% for 2013, creating an expectation of compound annual growth of 6.13% from 2010 through to 2013. By way of comparison, the SIA says that total sales for 2008 amounted to US$248.6 billion, down slightly from the US$255.6 billion figure for 2007 (−2.8%). In 2009, sales fell once again, down by 9% to US$226.3 billion.
Commenting on the 2011 forecast, SIA president Brian Toohey pointed to the importance of fostering innovation if the sector is to thrive. “At a time of economic uncertainty, the semiconductor industry continues to be a bright spot in the US economy. Now more than ever, it’s clear that the only way to spur lasting economic growth is to empower the private sector to do what they do best—innovate, compete and grow. These projections are simply estimates; it’s important to remember that US policy can positively or negatively affect these growth numbers. This is why we are urging the Congress and Administration to move now to enact pro-business, pro-growth policies that will lead to a rebirth of the innovation economy.”
The SIA says that all regional markets, namely the Americas, Europe, and Asia, with the exception of Japan, are projecting year-on-year growth for 2011. Japan’s contraction, it points out, is understandable in the light of the impact of the March 2011 earthquake and tsunami, with some of the growth in other regions increased as a result of supply-chain shifts that helped to alleviate disruption.
Additionally, double-digit growth in 2011 and 2012 is projected for sensor and micro I/C categories. Rising demand for consumer electronic devices and safety-related mandates specifically in the automotive sector are driving the growth in the sensor category. Micro I/Cs are projected to grow 11.7% year-on-year in 2011 and 13.1% in 2012, in part driven by the rapid increase in the tablet segment, as well as increased smartphone usage, especially in China, India, and other emerging markets.
The Asian Semiconductor Industry
The Asian semiconductor industry began in the 1960s, with small pockets of foreign investment. Japan established a semiconductor industry in the 1970s, and in 1979 Fujitsu became the first company to mass-produce 64 KB memory chips, with Japan cornering the world market in memory chips by the mid-1980s, passing the United States in semiconductor production volumes. By the late 1980s, South Korea had developed a thriving memory-chip industry, with companies such as Samsung enjoying rapid growth. Taiwanese investment generated the world’s first “on-demand, fab-for-hire” plant, and Singapore, Malaysia, and China have all developed significant chip industries.
According to Colonel Howe’s study, the first US outsourcing investment in the sector was by Fairchild Semiconductor in Hong Kong in 1961. Outsourcing began with chip assembly (bringing the various components together to complete a printed circuit board), then moved to fabrication, and, later, in the 1980s, to chip design.
Asia now accounts for around 60% of global sales, and is home to many of the world’s leading chipmakers. According to SEMI, the global industry association serving the manufacturing supply chain for the micro and nano-electronics industries, investment in fab equipment in Southeast Asia grew by 108% through 2010, and SEMI is projecting 36% growth for 2011. “The sector is experiencing solid growth, with Southeast Asia contributing some 7% of global semiconductor fab capacity in 2010 at the 300 mm die level. The sector as a whole grew 21% in 2010.” The Singapore Economic Development Board says that electronics output grew by 60% through the first half of 2010, riding the recovery in global demand and boosted by capacity expansions by Singapore electronics companies. Singapore is the second-largest center of semiconductor output and manufacturing in Asia. Other regions in Southeast Asia reported similarly strong gains. According to Asia Electronics News, Taiwan is expected to emerge as the world’s largest semiconductor materials market in 2011. Total sales of semiconductor materials in Taiwan in 2010 were up 33% year on year, at US$9.11 billion.
Following the downturn seen in 2008 and 2009, Asian chipmakers have been experiencing something of a rebound in demand. Better-than-expected demand for personal computers and limited output by smaller players have boosted prices of dynamic random access memory (DRAM) and NAND memory chips, which began to benefit sector leaders such as Samsung Electronics, Hynix Semiconductor, and Toshiba towards the end of 2009, and this trend continued into 2010. However, the sector’s history is marked with volatile cycles of shortages and oversupply, and some analysts have warned the industry could be setting itself up for a supply glut. Memory chipmakers returned to profit in late 2009 after several quarters of losses, as prices rebounded and PC demand grew, aided by strong growth from China.
According to the Electronic Industries Alliance (EIA), a national trade organization that includes the full spectrum of US electronics manufacturers, national paranoia about where semiconductor work is done, or who “owns” what, has the potential to be hugely counterproductive to the sector. As a sector that thrives on R&D and innovation, the greater the number of participants, and the freer and more open the semiconductor “universe” is, the better it will be for the sector, and for the global economy, the EIA argues.
In an initiative titled “The Technology Industry at an Innovation Crossroads,” the EIA argues strongly against what it sees as the possibility of protectionist policies stifling innovation in the sector. “The core value of a knowledge-based company or society should be innovation,” it says. The United States and other Western countries should not be worrying about the possibility of China or India dominating the semiconductor industry in years to come; they should be working to develop a vision and a strategy for the sector, the EIA says.
This opens up one of the really important themes for the sector. Electronics has enabled the globalization of business, and the electronics sector has itself, in turn, been shaped by globalization. “The resulting reorganization of manufacturing (in all sectors) along global lines, plus the creation of new, globally competitive service and knowledge-based industries (of which the semiconductor industry is a prime example) poses unprecedented challenges…” the EIA says.
Countries with weak science and mathematics education, and with a dearth of R&D funding, will fall behind countries that prioritize these areas. Similarly, countries that stay open to what the EIA calls “the brightest foreign minds,” and that allow their companies to recruit the best from around the world, will prosper, while those that seek to restrict highly paid knowledge jobs to their own nationals will fall behind.
The semiconductor industry has already enabled software designers to go some way down the road to “virtualizing” our world, creating immensely powerful tools for solving real-world problems far faster than ever before, and enabling new products, new drugs, and new forms of entertainment (of which the video-games console and the mobile phone are two stunning examples) to be brought to market extremely rapidly.
It is a safe bet that this sector is going to change life as we know it almost beyond recognition over the coming decades. As such, it is certainly a sector worth watching.