Texas Instruments NASDAQ:TXN is approaching the final stages of a six-year elevated capital expenditure (CapEx) cycle focused on building 300mm wafer fabrication plants. In a February 2025 management update, the company stated it was “nearly 70% through” this cycle. As of late December 2025, the cycle is approximately 83% complete, with FY 2025 CapEx expected around US$5 billion and FY 2026 guidance between US$2–5 billion. The elevated spending is set to wind down by the end of 2026.
Once complete, Texas Instruments expects to generate meaningful free cash flow per share growth across various market conditions. Given the company’s long-standing policy of returning all free cash flow to shareholders through dividends and share repurchases, the conclusion of this investment phase, in my opinion, bodes well for future returns to shareholders.
A Snapshot of Texas Instrument’s CapEx History
From 2018–2020, annual CapEx ranged between US$649 million and US$1.1 billion. The current elevated cycle began in 2021 and represents a major step-up in spending to expand advanced analog manufacturing capacity, particularly on larger 300mm wafers.
Texas Instruments – A Brief History
Founded in 1930 as Geophysical Service Incorporated, TXN initially focused on seismic exploration for the oil industry. It contributed to WWII defense electronics before renaming to Texas Instruments in 1951. The company made history in 1958 when engineer Jack Kilby invented the first working integrated circuit—a germanium-based device that combined multiple components on a single chip, laying the foundation for modern electronics.
Texas Instruments also produced the first silicon transistor (1954) and the world’s first handheld calculator (1967). In the early 1970s, TXN and Intel NASDAQ:INTC both worked on single-chip microprocessor designs for Datapoint. While Intel commercialized the 4004 (1971), TXN’s design (TMX 1795) was never released. TXN subsequently pivoted away from general-purpose microprocessors toward specialized microcontrollers (e.g., TMS1000 series) and consumer products like calculators.
The 1980s–1990s proved challenging. Intense competition from Japanese firms in DRAM (memory) and consumer electronics eroded margins, leading to losses, write-offs (notably the TI-99/4A home computer), and restructuring. TXN was seen as an unfocused conglomerate.
Strategic Pivot to Analog Semiconductors
In the late 1990s, Texas Instruments decisively shifted toward analog semiconductors. Key moves included:
- Divesting DRAM operations to Micron (1998)
- Acquiring high-performance analog specialist Burr-Brown (2000, $7.6 billion)
- Exiting wireless baseband (2008–2012)
- Acquiring National Semiconductor (2011, $6.5 billion)
These steps transformed TXN into the world’s largest analog chip maker. As of FY 2024, 78% of revenue comes from analog products.
This focus has driven strong returns on invested capital. Cash Flow Return on Investment (CFROI) consistently exceeded 30% from 2017–2022, reflecting a durable competitive advantage. The recent dip in returns (2023–2024) is largely attributable to the heavy CapEx phase, as newly constructed assets are not yet fully productive.
To help understand Texas Instruments’ competitive advantage it’s necessary to delve a little deeper into analog semiconductors. Let’s start with a basic comparison of digital and analog semiconductors.
Digital vs. Analog Semiconductors: Key Differences
Digital signals are discrete (binary 1s and 0s), making them highly noise-resistant, scalable, and ideal for computing, storage, and communication. This enabled decades of aggressive scaling under Moore’s Law.
Analog signals are continuous, mirroring real-world phenomena (temperature, sound, pressure, light). Analog chips excel at interfacing with the physical environment but are more susceptible to noise and degradation.
Why Moore’s Law Applies to Digital, Not Analog
In his 1965 paper, Gordon Moore predicted that digital circuits would scale dramatically with shrinking transistors, while analog (“linear”) circuits would see limited benefit due to physical constraints—particularly the difficulty of integrating large-value capacitors and inductors on-chip. History has proven him correct.
Digital chips pursued ever-smaller process nodes (now at 2nm-class), delivering higher speed and efficiency. Analog designs, however, face serious challenges at smaller nodes, especially transistor matching in circuits like operational amplifiers (op-amps). Tiny manufacturing variations cause mismatches that degrade performance. As a result, most analog production remains on mature nodes (45nm–130nm), introduced in the early-to-mid 2000s.
The Wafer Size Revolution: 200mm vs. 300mm
Larger wafers improve economies of scale. 300mm wafers provide ~2.25× more surface area than 200mm, enabling 30–40% lower cost per die in high-volume production.
Digital manufacturers (logic, DRAM, NAND) adopted 300mm wafers rapidly from the early 2000s. Analog stayed on 200mm longer due to lower volumes, long product lifecycles (especially in automotive/industrial), and technical hurdles—particularly achieving uniform temperature and defect control across the larger wafer, which is critical for precise transistor matching.
Texas Instruments’ 300mm Breakthrough
TXN capitalized on a rare opportunity in 2009. After Qimonda’s bankruptcy, TI acquired advanced 300mm equipment from its Virginia fab for just $172.5 million—a fraction of new-build cost. This enabled TI to convert its RFAB facility in Richardson, Texas, into the world’s first dedicated 300mm analog fab and gain experience with 300mm wafer technology.
Today, TXN is expanding 300mm capacity primarily in North America (Richardson, Lehi, Sherman). On December 17, 2025, TXN announced first production at SM1 in Sherman, Texas—a major milestone in realizing lower-cost analog production.
SM1 & SM2 fabs in Sherman, Texas. Source: Texas Instruments
Their strategy of moving production from 200mm to 300mm wafer manufacturing is being realised and providing a path to significantly cheaper products. Here is a presentation slide from Texas Instruments:
Source: Texas Instruments
Conclusion
Since its late-1990s pivot, Texas Instruments has developed deep expertise in analog semiconductors, delivering consistently high returns indicative of a lasting competitive advantage. Analog design remains a specialized skill requiring years of experience—difficult to replicate and supporting enduring profitability.
By investing heavily in 300mm wafer fabs, TXN is positioning itself to produce analog chips 30–40% cheaper than on 200mm, reinforcing its market leadership. As the six-year CapEx cycle concludes in 2026, shareholders can look forward to significantly improved free cash flow generation and enhanced returns through dividends and buybacks.
Important Disclosure: Roger Breuer, CFA, ACCA is an investment analyst for the Atlas Global Equity Income Fund, which maintains a position in Texas Instruments shares. The views expressed in this article are for informational purposes only and should not be construed as investment advice, financial recommendations, or an offer to buy or sell any securities.
Comments (0)