Nvidia Gets All the Credit, but These 4 Stocks Are Quietly Capturing the $725 Billion AI Buildout

The article argues that AI infrastructure spending, estimated at $725 billion this year, is creating strong earnings growth opportunities beyond Nvidia. Taiwan Semiconductor, Broadcom, ASML, and Arm are highlighted as beneficiaries, with expected earnings growth of 22%, 49%, nearly 30%, and almost 25% annually, respectively, over the next three to five years. The piece is broadly bullish on the AI semiconductor supply chain, though it is framed as commentary rather than new company-specific news.

Analysis

The key implication is that the AI monetization stack is broadening from the visible chip layer into the “pick-and-shovel” layers with the strongest pricing power. That favors the companies with structural bottlenecks — foundry capacity, lithography tools, and custom silicon design talent — while commoditizing more of the incremental value for generic hardware assemblers and lower-tier component vendors. The second-order effect is that AI capex is becoming less of a single-name Nvidia trade and more of a multi-year supply-chain capacity trade, which should keep earnings revisions positive even if GPU unit growth normalizes.

Broadcom is the clearest beneficiary of hyperscalers’ desire to reduce dependence on one supplier; custom accelerators tend to win when customers care more about watts-per-token and total cost of ownership than brand. That creates a subtle threat to Nvidia’s long-duration share premium, not necessarily near-term revenue, but the market’s assumption that every AI dollar must flow to GPUs. Meanwhile, TSM and ASML are leveraged to the same secular build-out, but their risk is less demand and more execution: any delayed wafer-ramp, export restriction, or capex pause can hit the order book with a lag of 2-4 quarters.

The contrarian point is that consensus still treats AI infrastructure as one homogeneous winner, when the next leg likely comes from bottlenecks and substitution. If hyperscalers keep pushing custom silicon, the relative winner could be AVGO/TSM/ASML over NVDA on an earnings-revision basis, while Intel remains mostly a relative loser unless it can meaningfully close process gaps. Arm is interesting as an embedded royalty stream with upside from AI CPUs, but the market may be underestimating how long it takes for that to become material versus the already-visible foundry and tools cycle.

Key Decisions for Investors

• Long AVGO vs NVDA on a 6-12 month horizon: AVGO has the cleaner re-rating path from custom AI silicon monetization; use a 1:1 dollar pair with a stop if NVDA re-accelerates server backlog commentary.
• Add TSM on weakness for a 12-24 month hold: this is the highest-quality capacity bottleneck exposure; best risk/reward is on any 5-8% pullback tied to Taiwan headline noise rather than fundamentals.
• Own ASML for the next 12 months as a delayed-beneficiary capex play: entry is attractive on any export-control-driven drawdown; risk/reward is asymmetric because order backlog should lag but not break unless AI capex rolls over broadly.
• Initiate a small long ARM position as a call on AI CPU adoption, but size it below AVGO/TSM: better viewed as a 2-3 year optionality trade than a near-term earnings compounder.
• Hedge the basket with a small short INTC against long TSM/ASML: if AI spend remains concentrated in leading-edge manufacturing, Intel’s catch-up story is the most vulnerable to prolonged underinvestment and share loss.