The Next Chip Crunch: How AI Demand Is Tightening Analog Supply and Reshaping the Semiconductor Trade

Author: Zion Zhao Real Estate | 88844623 | 狮家社小赵 | wa.me/6588844623

Author’s note and disclaimer: For general education and market literacy only. Not financial, investment, legal, accounting, or tax advice, and not an offer, solicitation, or recommendation. Information is general and may be inaccurate or change. No liability accepted. Investing involves risk, including loss of principal; past performance is not indicative of future results. 

Beyond GPUs: Why Mature-Node and Analog Chips Could Become the Market’s Next Bottleneck

The Next Semiconductor Bottleneck May Not Be AI Compute. It May Be the Foundational Chips Behind the Physical Economy

Wall Street is still fixated on the obvious semiconductor story: artificial intelligence is absorbing extraordinary amounts of capital, talent, and manufacturing capacity. That view is not wrong. Demand for advanced logic, high bandwidth memory, and advanced packaging remains intense, with companies such as TSMC and Micron continuing to report strong AI-driven momentum (Micron Technology, 2024, 2025; TSMC, 2025a, 2026). But markets often misprice what they are not watching. The next semiconductor constraint may not come from the glamorous edge of AI. It may come from the mature-node, analog, mixed-signal, power, and embedded chips that connect intelligence to the real world.

This is where the real debate should begin. The issue is not that the industry has suddenly stopped caring about foundational semiconductors. It is that AI has become so profitable and strategically urgent that capital allocation is increasingly tilting toward leading-edge nodes and advanced packaging. TSMC’s disclosures make that clear. For 2026, the company stated that the large majority of its capital spending would go toward advanced process technologies, with smaller portions directed to specialty technologies and advanced packaging, testing, mask making, and related areas (TSMC, 2026). That is rational from a shareholder and market-demand perspective. But it also means mature-node supply may become more fragile at the margin if demand returns more quickly than inventories, capacity additions, and qualification cycles can respond.

That matters because so-called trailing-edge chips are not obsolete chips. They are the semiconductors that quietly run the physical economy. They sit inside cars, factory systems, robotics, power electronics, industrial equipment, medical devices, aerospace platforms, network hardware, and consumer electronics. The U.S. Bureau of Industry and Security has already highlighted the strategic dependence of major industries on mature-node semiconductors, including automotive, aerospace and defense, industrial equipment, healthcare, and consumer products (Bureau of Industry and Security, 2024). Texas Instruments similarly reported that about 95 percent of its 2025 revenue came from analog and embedded processing semiconductors, with roughly 75 percent tied to industrial, automotive, and data center markets (Texas Instruments Incorporated, 2026). These are not peripheral chips. They are foundational.

The popular mistake is to think that AI demand ends with hyperscalers and GPUs. It does not. As AI diffuses into physical systems, demand broadens. TSMC itself has described a future in which AI reaches beyond data centers into smartphones, automobiles, personal computers, and Internet-of-Things devices (TSMC, 2025a, 2026). That means more sensors, power management, signal chains, connectivity, controllers, and embedded processing. In short, more foundational semiconductors.

This is where the current setup becomes strategically interesting. Several end markets that matter to analog and embedded chip suppliers, especially automotive and industrial, have not yet returned to full strength. NXP reported that 2024 automotive revenue fell 4.4 percent and industrial and IoT revenue fell 3.5 percent year over year (NXP Semiconductors N.V., 2025). Yet Analog Devices later signaled that bookings had improved and customer inventory digestion was ending, suggesting recovery rather than collapse (Analog Devices, Inc., 2025a, 2025b). TSMC also noted that non-AI markets had bottomed out and were experiencing a milder recovery by late 2025 (TSMC, 2026). This is the kind of environment that creates future stress: demand is not booming today, so the market stays complacent. But inventories are lean, product qualification cycles are long, and the broader industry is allocating attention elsewhere.

That is why Texas Instruments stands out. Not because it is the only beneficiary, and not because a shortage is guaranteed, but because it has spent years building exactly the sort of manufacturing posture that becomes valuable when customers prioritize supply assurance. TI has invested heavily in internal manufacturing, especially 300 millimeter wafer capacity, while many peers remain more reliant on hybrid or external foundry models. The company says it aims to source more than 95 percent of its wafers internally by 2030, with more than 80 percent on 300 millimeter wafers, which materially lowers cost relative to 200 millimeter production (Texas Instruments Incorporated, 2026). It has also maintained an inventory strategy designed to support customers through volatility rather than optimize optics quarter by quarter.

That is a serious strategic distinction. In semiconductors, particularly analog and embedded processing, resilience is not just about having product. It is about having qualified, trusted, regionally dependable, cost-competitive supply. TI argues that owning its manufacturing, process, and packaging technologies gives it better cost control, better customer service, and stronger geopolitical dependability (Texas Instruments Incorporated, 2026). If the next phase of AI is physical AI, then those advantages become more meaningful, not less.

Still, a credible thesis needs balance. The future is not a simple story of universal shortage. The BIS has warned that China is expected to account for a very large share of new mature-node capacity, and that this could create pricing pressure in some categories such as analog, discrete, mixed-signal, and power semiconductors (Bureau of Industry and Security, 2024). That means the real opportunity may not be a blanket shortage across all mature nodes. It may be a more selective premium on trusted, high-quality, regionally aligned supply for industrial, automotive, and strategic applications where qualification barriers and customer relationships matter most.

That is the point investors and industry leaders should focus on. The AI trade is not only about the chips that train large models. It is also about the less celebrated semiconductors that allow intelligence to move into cars, machines, robots, healthcare devices, and critical infrastructure. The market has spent years rewarding the obvious winners in advanced compute. The next winners may be the companies that quietly spent through the downcycle, built internal capacity, protected supply, and stayed close to industrial reality.

The semiconductor story is no longer just about faster computing. It is increasingly about who can reliably supply the physical world that AI is now beginning to transform.

References

Analog Devices, Inc. (2025a, November 25). Analog Devices reports strong fourth quarter and fiscal 2025 financial results.

Analog Devices, Inc. (2025b, May 29). Analog Devices Inc at Bernstein Strategic Decisions Conference on May 29, 2025.

Bureau of Industry and Security. (2024, December). Public report on the use of mature-node semiconductors.

Micron Technology. (2024, December 18). Financial results.

Micron Technology. (2025, March 20). Financial results.

NXP Semiconductors N.V. (2025, February 20). Form 10-K for the fiscal year ended December 31, 2024.

Taiwan Semiconductor Manufacturing Company. (2025a). 2024 annual report.

Taiwan Semiconductor Manufacturing Company. (2026, January 15). Q4 2025 Taiwan Semiconductor Manufacturing Co Ltd earnings call (Chinese, English) on January 15, 2026.

Texas Instruments Incorporated. (2026). 2025 annual report.

The Hidden Semiconductor Shortage: TSMC, AI Capacity Shifts, and the Rising Strategic Value of Analog Chips

Artificial intelligence is monopolising semiconductor attention, but the next constraint may emerge in mature-node analog and embedded chips powering cars, factories, robotics, and infrastructure. As capacity tilts toward leading-edge demand, Texas Instruments appears unusually well positioned to benefit from any future supply tightening.

This matters to Singapore property clients because semiconductor cycles do not stay confined to stock markets. They shape employment, business expansion, capital flows, industrial demand, and ultimately housing sentiment. When artificial intelligence and advanced manufacturing attract more investment, the effects can ripple through office demand, expatriate leasing, housing budgets, and long term confidence in innovation-led economies such as Singapore. For buyers, this strengthens the case for owning quality property in locations supported by jobs, infrastructure, and economic resilience. For sellers, it sharpens the importance of timing, positioning, and pricing in a market increasingly influenced by global capital and technology wealth. For landlords, it highlights how tenant demand can be supported by professionals and firms linked to growth sectors. For investors, it reinforces why property should be assessed not only by headline prices, but also by structural trends in the global economy.

In a market as competitive and nuanced as Singapore, clients need more than a transactional agent. They need a strategist who can connect macroeconomics, capital markets, policy, and real estate fundamentals into clear, actionable advice.

If you are planning to buy, sell, rent, or invest in Singapore property, engage me for a professional, data-driven, and globally informed assessment. I help clients cut through noise, identify real opportunities, manage risk, and make confident property decisions grounded in both local market knowledge and broader economic reality.