Battle Lines Are Drawn in the Global Chip Industry

Fab versus fabless, and the meaning of the production node.

The spread of microprocessors as embedded components in multiple industrial and consumer goods dramatically increased demand for these vital devices.  

Currently, a small number of companies and nations effectively control the global supply of high-performance integrated circuits, but this may be changing. Recent legislation in the United States and China suggest a future where regional economic blocs regulate the manufacturer of integrated circuits as a strategic resource.  

In the meantime, continued supply chain problems hamper production of many consumer durables such as automobiles. With the time-to-market of a new fab facility being on the order of two years, only falling demand can address shortages in the short term.

Access all episodes of This Week in Engineering on engineering.com TV along with all of our other series.

Transcript of this week’s show:

To see any graphs, charts, graphics, images, and/or videos to which the transcript may be referring, watch the above video.

Every age can be defined in engineering terms by the critical materials they relied on. The 19th century was an age of steel, the essential technology for large buildings, bridges, railroads and eventually automobiles. The 20th century was defined by polymers and by copper, the critical technology that enabled the electrification of the world and the birth of mass consumer society.  

So far in the 21st century, the critical material is silicon—or more specifically, integrated circuits. ICs have evolved from simple logic gates, op amps and timers to modern systems on a chip, in form factors small enough and at costs low enough to allow their integration into everything from nuclear weapons to smartphone charging cables.   
 
Those that control the supply of integrated circuits have considerable control over the global economy. That kind of control has inevitable political consequences, and with the serious disruption in semiconductor supplies due to COVID-19, the post-COVID global semiconductor environment will define economic growth for the next several decades.  
 
Pre-COVID, the industry was dominated by Taiwan, accounting for over 100 billion dollars in annual sales, with a single company, the Taiwan Semiconductor Manufacturing Company (TSMC) accounting for over 50 percent of the global market in advanced processors.  
 
TSMC and Korea’s Samsung have near control over the global advanced chip market. With advanced semiconductors needed for almost every consumer, commercial, industrial and military application, control of integrated circuits has a powerful geopolitical dimension.   
 
The Biden Administration’s new CHIPs and Science bill launches a 10-year, 40-billion-dollar program for semiconductor manufacturing on American soil. Concurrent with the bill, American producers Qualcomm and GlobalFoundries announced a 4-billion-dollar investment to produce semiconductors domestically.  

These investments are huge, but from an engineering perspective there are two critical takeaways: one is that who manufactures the semiconductors is important, and the other is in the size of the semiconductor gates.  
 
The first factor is important. Starting the early 80s, excess manufacturing capacity in semiconductors created a contract manufacturing industry, allowing a generation of integrated circuit design firms to focus on IP and outsource the actual chip production. According to market research firm TrendForce, half of the world’s IC revenue from fabless companies comes from three firms: Qualcomm, Broadcom and Nvidia. 

Heavyweights such as Samsung and TSMC not only design integrated circuits, but manufacture them as well, primarily in Asia. Both United States and the People’s Republic of China have implemented major policy changes to drive investment in domestic foundry capability to reduce reliance on foreign sources.  

Taiwan, in the dominant global supply position, is important enough to Western economies that the country’s foundry base represents a “silicon shield”—meaning Western nations are at risk should the nation be militarily overrun by the PRC. Development of domestic capabilities are a significant national security issue for China and the U.S., but the second factor that is essential is the size of the MOSFET semiconductor gates or junctions that defines processor power and memory capacity.  

Historically this was gauged by physical size, in nanometers, and referred to as the “technology node” or “process node.” Smaller numbers mean greater circuit density, although today, process node numbers refer to a generation of chips and not a specific gate length or pitch. This makes it difficult to directly compare semiconductors made by different companies based on process node nomenclature.  

From a manufacturing engineering perspective, demanding devices such as smartphones and laptops require faster, more power-efficient process nodes such as 22, 16, 14 and 10 nanometers, and single-digit nodes are entering production now.  But a huge volume of chips is consumed by industries that function well with trailing edge technology such as 45 or 90 nanometers, especially the automotive industry.  
 
For automakers, older technology chips are perfectly adequate for most body and chassis control functions such as climate control and power assists, and are cost-effective. But foundries that are switching to advanced technology nodes to serve the high-tech markets can yield more devices per silicon wafer at the smaller node size, and are reluctant to continue to invest in older, lower-technology factories.  
 
In the meantime, the chip shortage continues. According to IHS Markit, the booking lead-time for automotive manufacturers for dedicated semiconductors has stretched from 12 weeks to a year, and the research firm reports that several OEMs are exploring bookings at larger node Tier 2 foundries over a year in advance. The chips are not only mission-critical in automaking, but they also represent an increasing portion of production cost. According to Intel, in 2019, four percent of the bill of materials of the car was represented by semiconductors. The company expects that to grow to 15 percent in 2025, and 20 percent by 2030. The market for automotive chips alone will grow from 50 billion dollars to 115 billion dollars in 2030.  
 
The market is huge, and the investment decisions and political maneuverings of major manufacturing nations in Asia, America and Europe may decide where the planet’s mass consumer goods are manufactured in the middle of the century. 

Written by

James Anderton

Jim Anderton is the Director of Content for ENGINEERING.com. Mr. Anderton was formerly editor of Canadian Metalworking Magazine and has contributed to a wide range of print and on-line publications, including Design Engineering, Canadian Plastics, Service Station and Garage Management, Autovision, and the National Post. He also brings prior industry experience in quality and part design for a Tier One automotive supplier.