TSMC Positioned as Better AI Play Than Nvidia at Cheaper Valuation

TSMC profits from AI across multiple fronts while Nvidia remains concentrated in data centers.
TSMC manufactures chips for data center AI, consumer PCs, and smartphones, while Nvidia's opportunity is primarily in enterprise infrastructure.
Mark

So the argument here is basically that TSMC is cheaper and does the same work as Nvidia. Is that really the comparison?

Mimi

Not quite. TSMC doesn't design chips—it manufactures them. Nvidia designs the chips and TSMC builds them. The point is that TSMC gets paid by Nvidia, AMD, Intel, Apple, Qualcomm, and others. It's not betting on one company winning the AI race.

Luke

But that's also why TSMC's growth is capped differently. Nvidia's growth depends on how much AI infrastructure gets built. TSMC's growth depends on how much of that infrastructure gets built on TSMC's fabs versus competitors' fabs. That's a different risk.

Mimi

True, but TSMC has about 54 percent of the global foundry market. And the source shows AMD, Intel, and Nvidia are all using TSMC. That's pretty concentrated customer power.

Mark

What about the valuation numbers? 31 times earnings versus 71 times—is that a real discount or just because Nvidia is growing faster?

Luke

Nvidia's forward multiple is 45 times, and TSMC's is 26 times. So even accounting for Nvidia's faster growth, TSMC is cheaper. But the source doesn't tell us whether that gap is justified by growth rates or whether it's a genuine mispricing.

Mimi

The source projects TSMC will grow revenue 23 percent in 2024 and 20-plus percent in 2025. Nvidia is expected to almost double. So some of that valuation gap makes sense.

Mark

What about the consumer device angle—iPhones, PCs, Snapdragon chips? How real is that opportunity?

Mimi

Gartner is forecasting 295 million AI-capable phones and PCs shipped in 2024, up from 29 million the year before. But only 22 percent of all devices shipped this year are expected to be AI-capable, so there's room to grow.

Luke

That's a forecast, though. We don't know if those devices will actually drive meaningful revenue for TSMC or if the margins will be thin. The source doesn't break down revenue by product type.

Mark

So TSMC is less risky than Nvidia because it's diversified, but also less explosive?

Mimi

That's one way to frame it. It's the foundry play rather than the design play. You get exposure to AI growth without betting everything on Nvidia's continued dominance.

Luke

And you get it at a cheaper valuation. But you're also getting a company that's more dependent on geopolitical risk—Taiwan's relationship with China matters a lot more for TSMC than for Nvidia.

  • Nvidia's 727% stock surge in eighteen months has compressed its valuation to 71x trailing earnings, raising the question of how much future growth is already priced in.
  • TSMC sits at the manufacturing center of the AI race — producing chips for Nvidia, AMD, Intel, Apple, and Qualcomm simultaneously — yet trades at just 31x trailing earnings, a significant discount for a company with no single point of failure.
  • Consumer AI is accelerating the urgency: Gartner projects AI-capable PC and smartphone shipments could leap from 29 million to 295 million units in a single year, and TSMC holds production contracts across nearly all of them.
  • After a 9% revenue decline in 2023, TSMC's momentum has sharply reversed — April 2024 revenue surged nearly 60% year over year, signaling that the semiconductor cycle has turned decisively in its favor.
  • Analysts project 23% revenue growth in 2024 and 20%+ in 2025, positioning TSMC as a compounding beneficiary of AI adoption across data centers, personal computers, and smartphones alike.

As artificial intelligence reshapes the global economy, investors have flocked to the most visible names — yet the deeper infrastructure enabling that transformation often carries less of the speculative premium. Taiwan Semiconductor Manufacturing Company, the foundry that physically produces chips for Nvidia, AMD, Apple, and nearly every other major player, trades at roughly half the earnings multiple of its most celebrated customer. In the long arc of technological revolutions, those who build the roads sometimes outlast those who first race upon them.

Nvidia's ascent has been one of the most dramatic in recent market history — a 727 percent stock gain in under eighteen months, driven by voracious demand for AI chips. Revenue jumped 126 percent in fiscal 2024, and analysts expect the company to nearly double again to $120 billion in the current year. For investors who caught the wave early, the returns have been extraordinary. For those who did not, the valuation — 71 times trailing earnings — raises a harder question.

There is a quieter company at the center of the same story. Taiwan Semiconductor Manufacturing Company does not design the chips that power AI; it manufactures them. Nvidia is its second-largest customer, accounting for 11 percent of 2023 revenue. But TSMC also produces AMD's MI325X AI chip, Intel's Lunar Lake processors for AI-enabled PCs, Apple's forthcoming 2-nanometer iPhone components, and Qualcomm's Snapdragon smartphone processors. In a race where the winner among chip designers remains uncertain, TSMC profits from all of them.

The consumer dimension of AI adds another layer. Gartner forecasts that AI-capable PC and smartphone shipments could surge from 29 million units in 2023 to 295 million in 2024 — and even then, only 22 percent of all devices shipped would carry AI features, leaving an enormous runway ahead. TSMC holds manufacturing contracts across this entire landscape, giving it exposure that no single chip designer can match.

Financially, the company's trajectory has turned sharply upward after a difficult 2023. First-quarter 2024 revenue rose 13 percent year over year, and April revenue accelerated to nearly 60 percent growth compared to the same month last year. Analysts project roughly 23 percent revenue growth for the full year, followed by 20-plus percent in 2025. At 31 times trailing earnings — less than half Nvidia's multiple — TSMC offers a different kind of bet on the AI era: not on who wins the design race, but on the infrastructure that makes the race possible at all.

Nvidia's rise has been staggering. Since the start of 2023, the graphics card maker's stock has climbed 727 percent in less than eighteen months, propelled by insatiable demand for the chips that power artificial intelligence systems. The numbers reflect this momentum: revenue in the fiscal year ending January 2024 jumped 126 percent to $60.9 billion, while adjusted earnings per share surged 288 percent to $12.96. Analysts expect the company to nearly double revenue again in the current fiscal year, reaching just over $120 billion, with earnings per share more than doubling to $27.03. Nvidia's dominance in AI chip design has made it the obvious play for investors betting on the technology's explosive growth.

But there is another semiconductor company positioned to capture the AI opportunity with less of the valuation premium already baked in. Taiwan Semiconductor Manufacturing Company, or TSMC, trades at 31 times trailing earnings compared to Nvidia's 71 times. Even on a forward basis—where Nvidia's multiple compresses to 45 times earnings—TSMC comes in cheaper at 26 times estimated earnings. The valuation gap matters because TSMC manufactures the very chips that Nvidia designs. In 2023, Nvidia accounted for 11 percent of TSMC's revenue, making it the second-largest customer. But TSMC's reach extends far beyond Nvidia alone.

The Taiwan-based foundry is the manufacturing backbone for multiple chipmakers racing to compete in AI. Advanced Micro Devices is using TSMC's N5 and N6 process nodes to produce its MI325X AI chip, with plans to move its next-generation MI350X to TSMC's 3-nanometer process in 2025. Intel has contracted TSMC to manufacture Lunar Lake chips designed for AI-enabled personal computers. Nvidia itself relies on TSMC for its latest AI processors and has committed to using TSMC's 3-nanometer node for its Rubin chips, expected in 2026. This concentration of orders from the industry's heaviest hitters positions TSMC to benefit from the AI boom regardless of which chipmaker wins market share.

The opportunity extends beyond data centers into consumer devices. Apple is reportedly securing TSMC's 2-nanometer production capacity to embed AI features into iPhones and iPads. Qualcomm has tapped TSMC to manufacture chips for AI-enabled computers using its 4-nanometer process, and also relies on TSMC for its AI-focused Snapdragon 8 Gen 3 smartphone processors. Research firm Gartner forecasts that shipments of AI-capable personal computers and smartphones could explode from 29 million units in 2023 to 295 million units in 2024. Even at that scale, only 22 percent of all PCs and smartphones shipped this year are expected to have AI capabilities, suggesting a long runway of growth ahead. While Nvidia's AI opportunity remains concentrated in data centers, TSMC stands to profit from the technology's spread across multiple markets simultaneously.

TSMC's recent financial trajectory supports this thesis. The company's 2023 revenue fell 9 percent to $69.3 billion as the broader semiconductor market weakened, but conditions have shifted. First-quarter 2024 revenue climbed 13 percent year over year to $18.9 billion. April revenue jumped nearly 60 percent compared to the same month last year, a sharp acceleration from the 34 percent growth recorded in March. Analysts project TSMC's revenue will increase roughly 23 percent in 2024 to just over $85 billion, followed by 20-percent-plus growth in 2025. The company's exposure to multiple AI-driven segments—from data center chips to smartphone processors to PC components—suggests these growth rates could prove sustainable.

For investors who missed Nvidia's extraordinary run, TSMC offers exposure to the same secular shift in semiconductor demand at a more reasonable price. The global AI chip market is forecast to grow 40 percent annually through 2032 and generate more than $1.11 trillion in annual revenue. TSMC's role as the manufacturing partner for nearly every major chipmaker pursuing AI means the company is positioned to capture a meaningful share of that expansion. The question is not whether TSMC will benefit from AI adoption, but whether the market has already priced in that benefit as thoroughly as it has for Nvidia.

TSMC is playing a mission-critical role in helping Intel, AMD, and Nvidia produce the hardware necessary for AI training and inference.
— Investment analysis
Möchten Sie die ganze Geschichte? Das Original lesen bei Yahoo Finance ↗
Kontakt FAQ