Teradyne: A Deep Dive into the Quiet Giant of Semiconductor Testing

In a world obsessed with chip designers and AI accelerators, Teradyne (TER) occupies a quieter but essential corner of the semiconductor ecosystem — one that rarely makes headlines but underpins nearly every chip that ships. In a recent episode of Rebellious Allocations, host John Rotonti of Bastion Fiduciary sat down with Morningstar’s Will Kerwin to do a thorough business deep dive. Here’s what you need to know.

What Does Teradyne Actually Do?

Teradyne’s primary business — roughly 80% of its revenue — is semiconductor testing. At its core, the company builds the large, highly complex machines that validate chips after they’ve been manufactured. Before a GPU ships to a data center or a smartphone chip goes into an iPhone, it has to be tested to confirm the circuitry works correctly. That’s where Teradyne comes in.

The company performs two main types of testing: wafer test, which probes individual chips on a silicon wafer before they’re cut apart, and final package test, which tests fully packaged chips ready for shipping. The complexity of these machines is not to be underestimated — testing a cutting-edge mobile processor requires sending precise electrical signals in and verifying the exact signals that come back, all at high throughput. To illustrate how demanding this can be: a single Teradyne machine testing Apple’s latest iPhone chips can only test two chips at a time and takes up to two minutes per cycle. With roughly 200 million iPhones sold annually, the scale of this operation is staggering.

Beyond semiconductors, Teradyne has two additional business lines — wireless and networking testing, and an industrial robotics division — each contributing about 10% of sales.

A Practical Duopoly

The semiconductor testing market is about as concentrated as they come. Teradyne and Japanese rival Advantest together command more than 80–90% of the global market, making it a near-duopoly. Five years ago, the split was roughly 45–40 in Teradyne’s favor. Today, it has shifted significantly — closer to 55–30 in favor of Advantest.

The reason? End market exposure. Teradyne built its dominance on the back of the smartphone boom, developing a particularly close relationship with Apple for testing iPhone chips. Advantest, meanwhile, spent decades cultivating relationships with Nvidia and AMD. As the AI investment wave supercharged demand for GPUs, Advantest was perfectly positioned to benefit. Teradyne was not.

Kerwin is careful not to frame this as a management failure. “They were riding the hot hand of mobile for a long time, and it was hard to see just how explosive the growth would be for AI.” The more important question now is what happens next.

Teradyne’s AI Opportunity

The good news for Teradyne investors: the company is actively closing the gap. Kerwin expects Teradyne to qualify as a second-source supplier for Nvidia testing later in 2026. In the semiconductor supply chain, dual-sourcing is standard practice — even preferred suppliers typically see a 70–30 split with a secondary vendor, which benefits both buyers and vendors alike. Nvidia, known for its preference for multi-sourcing across its supply chain, is expected to follow that pattern.

Beyond Nvidia, Teradyne already has meaningful AI exposure through other channels. The company has strong relationships with SK Hynix and Samsung on the high-bandwidth memory side, and with Broadcom for its custom AI chips (known as TPUs or XPUs). The AI buildout isn’t a single chip story — it spans GPUs, memory, and custom silicon — and Teradyne has footholds across several of these areas.

The Robotics Business: Promising but Patient

Teradyne entered the robotics market through acquisitions in 2015 and 2018, building a business centered on two product categories: collaborative robot arms (cobots), which assist human workers on factory floors, and autonomous mobile robots, which move materials around warehouses. Neither category competes with humanoid robots — these are industrial tools built to work alongside people.

The robotics business grew steadily through its first five years but has stagnated since COVID. Weak industrial demand in Europe — where the business is heavily concentrated — has been the primary culprit. The segment currently operates near breakeven and represents about 10% of sales and roughly 0% of profit.

Kerwin remains cautiously optimistic about a recovery. He sees the long-term thesis supported by industrial automation trends, aging workforces, labor shortages, and onshoring activity in both the US and Europe. Teradyne’s key partnerships — including a relationship with Amazon for warehouse automation and a presence at CES showcasing cobots alongside Schneider Electric — suggest the technology is competitive, even if the market hasn’t rewarded it recently.

On the question of whether robotics belongs under the same corporate roof as semiconductor testing, Kerwin is measured: “I’m close to neutral, slightly positive.” The businesses share no synergies, but diversification has some merit given the cyclicality of chip testing and the concentration risk from Apple.

Competitive Moats and Pricing Power

Morningstar rates Teradyne as a wide-moat business — a designation it expects to hold for at least 20 years. The moat rests on two pillars: intangible assets and switching costs.

The intangible assets come from decades of R&D investment and the accumulated design expertise required to test essentially any chip in existence. Teradyne and Advantest together invest billions in R&D cumulatively — a level no new entrant could plausibly match. The switching cost element is subtler but real: customer relationships in this business involve shared roadmaps, long-term planning, and deep technical collaboration. Walking away from an established testing partner isn’t a simple procurement decision.

Pricing power works differently here than in most industries. Rather than annual price increases, Teradyne resets prices upward with each new product generation. As chips get more complex, older machines become bottlenecks — they still work, but they’re slower. The incentive for a customer like Apple to buy the newest Teradyne equipment for each new chip generation is straightforward: speed to market is worth far more than the cost of the machine.

Financials and Outlook

At roughly $50 billion in market cap and $50 billion in enterprise value (net cash of ~$165 million), Teradyne generated $3.2 billion in revenue in 2025, with a 20.4% GAAP EBIT margin and $450 million in free cash flow. Returns on equity run around 20%, and with CapEx at just 7% of sales, the business is capital-light relative to its complexity.

Management’s updated long-term financial targets call for 15–25% annual top-line growth, gross margins above 60%, and operating margins above 30%, translating to adjusted EPS of $9.50–$11 in roughly three to four years. Kerwin’s own model sits at the lower end — 18% annual revenue growth — but still projects roughly 30% annual EPS growth over five years when paired with meaningful margin expansion.

ROIC, which dipped during the 2022–2023 downturn, is expected to rebound to 25% in 2026, well above Morningstar’s estimated 8.5–9% cost of capital for the company. The balance sheet has moved from $1.3 billion in net cash in 2021 to $165 million today, largely due to the $450 million acquisition of Technoprobe (a wafer probing capability) and ongoing organic investment. Kerwin expects net cash to rebuild from here.

Key Risks to Watch

Kerwin flagged several risks worth keeping in mind:

Customer concentration. Apple and Broadcom together likely represent just above 10% of revenue each, while TSMC — which manufactures chips for both — may account for close to 25% of revenue. That’s meaningful concentration, though Kerwin notes it’s been moderated by growing compute exposure.

Cyclicality. This remains an inherently cyclical business. The 2022–2023 revenue declines of roughly 15% each year were part pull-forward normalization and part ordinary cycle. AI may be driving an unusually durable upcycle, but it doesn’t eliminate cyclicality.

AI dependency. As Teradyne leans further into AI-related testing, it stands to benefit substantially — but also becomes more exposed to any slowdown in AI capex. As Kerwin puts it, that’s “both a blessing and a curse.”

Advantest competition. The market share shift toward Advantest in recent years is real, even if Teradyne’s opportunity to recapture GPU-related business is now emerging.

The Bottom Line

Teradyne is a high-quality, capital-light industrial technology company with durable competitive advantages in one of the most technically demanding businesses in the semiconductor supply chain. After several years of being underexposed to the AI investment wave, the company appears positioned to meaningfully participate in GPU testing going forward, while its robotics business offers a potential long-term growth option as industrial automation demand recovers.