Access Official Technical PDF

Share Summary:LinkedInEmail

Semiconductor foundry competition is shaped by more than a company’s current market share. Process technology, the ability to manufacture at scale, customer demand and investment cycles all affect how foundries compete for business. A 2019 analysis of Taiwan Semiconductor Manufacturing Company (TSMC) illustrates how those factors can combine to create uncertainty—even for an established market leader.

The assessment was written around the technology and market conditions of 2018 and 2019. Its forecasts should be understood as period-specific rather than as a description of the market today. Its broader value is the way it connects manufacturing capability with commercial risk: a technology lead can attract customers, but demand timing and capital-intensive capacity decisions still matter.

Why semiconductor foundry competition matters

Foundries manufacture chips designed by other companies. Their competitive position depends partly on process nodes: successive generations of manufacturing technology intended to improve characteristics such as power use, performance and chip area. Moving to a new process can help a foundry win customer programs, but commercialization requires overcoming significant technical challenges.

In the period covered by the analysis, TSMC had a substantially larger foundry market share than Samsung Electronics. Yet the report argued that Samsung’s early commercialization of a 7nm process using extreme ultraviolet lithography (EUV) could give it a technical advantage. This distinction is important: market leadership and leadership at a particular process generation are not the same thing.

EUV lithography and the 7nm process race

The comparison centered on two approaches to 7nm production. TSMC had commercialized a first-generation 7nm process using deep ultraviolet (DUV) lithography, while Samsung had commercialized a 7nm process using EUV. TSMC was expected to introduce its own EUV-based process later, according to the analysis.

EUV uses shorter-wavelength light than DUV. In the report’s account, this could reduce the need for complex multi-patterning—the use of multiple patterning steps to create fine features—and could reduce mask requirements. The analysis also discussed potential benefits to design flexibility and manufacturing, including shorter fab cycle times and less variation. These are process-level considerations, not a guarantee that one supplier will win every customer or achieve a particular commercial outcome.

The report cited estimates that TSMC’s planned second-generation 7nm process could improve power efficiency, performance and area relative to its first-generation process. It also noted that side-by-side comparisons with Samsung were not available. That qualification matters: process labels alone do not establish equivalent capabilities, and comparisons depend on the specific implementation and evidence available.

Market share is not the only measure of strength

Samsung’s position was one part of a wider competitive picture. The analysis also identified China-based SMIC as a potential source of increased competition, while noting that the timing and technical difficulty of commercializing advanced EUV processes remained uncertain for several foundries. GlobalFoundries, meanwhile, had suspended work on its EUV-based 7nm node in 2018.

That decision had implications for customers as well as rivals. The report suggested that uncertainty around GlobalFoundries could create an opportunity for other foundries to pursue business, including potential 7nm production for AMD. For engineers and supply-chain decision-makers, the general lesson is that a supplier’s process roadmap and ability to execute can affect where customers consider placing manufacturing programs.

However, the report’s market-share figures also show why competitive claims need context. Samsung’s reported share rose in 2018, partly because production of its own Exynos mobile processors was included in its foundry sales. Changes in reported share can therefore reflect business structure and accounting scope as well as new customer wins.

Demand timing and investment cycles

Advanced manufacturing capacity only creates value when customer demand supports its use. The 2019 bear case argued that expected demand growth from autonomous vehicles and 5G could take time to become a major incremental driver of semiconductor sales. It also pointed to falling DRAM and NAND flash prices as a possible signal of softer conditions across the broader semiconductor sector.

The analysis connected these concerns to industry investment. It argued that heavy capital expenditure in 2017 could contribute to lower prices in 2019. This is a familiar challenge in capital-intensive manufacturing: companies invest in capacity ahead of expected demand, but if demand arrives later or proves weaker, utilization and pricing can come under pressure. The report also highlighted declining demand for cryptocurrency-mining machines as a risk to one customer segment.

For industrial automation leaders, these dynamics are relevant when assessing the resilience of semiconductor supply chains. A foundry’s technology roadmap matters, but so do customer concentration, the timing of end-market growth and the possibility that capacity investment outpaces near-term demand.

Customer exposure can amplify foundry risk

The source analysis also raised concerns about TSMC’s exposure to major customers, including Apple, and about the potential effect of competitive shifts on customer allocation. When a large customer changes its product plans or sourcing decisions, the consequences may extend beyond one chip program. For a foundry, that can influence sales expectations and investment priorities; for customers, it reinforces the importance of understanding supplier options and manufacturing roadmaps.

These are risks to evaluate, not outcomes to assume. The report presented them as reasons its 2019 earnings expectations differed from consensus, rather than as proof that TSMC would lose its leadership. It also acknowledged the company’s financial and technological strengths and its ability to respond in the next process generations.

Practical implications for manufacturing decision-makers

The 2019 case points to several questions worth asking when evaluating foundry capability and semiconductor supply exposure:

  • Which process is actually in production? Distinguish commercialized capability from announced plans or future targets.
  • What supports the technology comparison? Check whether performance, power and area claims are based on comparable data.
  • How diversified is demand? Consider exposure to major customers and end markets whose demand may shift.
  • Does capacity align with demand timing? Investment ahead of customer adoption can create pricing and utilization pressure.
  • How might competitors change customer options? Evaluate both established leaders and suppliers developing advanced processes.

Conclusion

Semiconductor foundry competition cannot be reduced to a single market-share figure or process-node announcement. The historical TSMC analysis shows how lithography choices, commercialization timelines, rival investment, customer exposure and demand cycles interact. For industrial automation and operations professionals, the practical takeaway is to assess semiconductor suppliers through both their manufacturing roadmaps and the market conditions that determine whether that capability can be sustained and accessed reliably.

Access Official Technical PDF

About The Author