This article is a summary of the subscriber piece published on July 31, 2026. Figures and price targets are available in the original.
In June’s back-end equipment deep dive, I argued that AI chip test was the binding bottleneck of this cycle, and that the field had grown while only one seed spot remained. A field, in my vocabulary, is a sector worth planting in; a seed is a position I actually take. That one seed was Advantest (6857 JP), and I pinned the verification event to its earnings release on July 29. That date has now come and gone. Let me say it upfront: the seed has sprouted. Today’s note is the scorecard on that call, and a promise to keep tracking this field.
The same week brought one more scene. On July 24, Nomura published a lengthy sector report covering the entire semiconductor test value chain, initiating six Taiwanese test names, all at Buy. Somewhere in the past few days, “test is the next bottleneck in AI chips” stopped being a contrarian frame and became the market’s common language.
Step back and my read is this. The binding constraint on AI infrastructure is, in the end, the sheer quantity of chips the industry can supply — the Q. And expanding Q is no longer a corporate capex question; it has become a national-level project. Fail to scale compute fast enough and you either fall behind or pay dearly for what you get. Test is the gate every chip must pass before it ships. When the gate narrows, the whole field’s harvest runs late.
Bigger chips, longer exams
Why test, why now? Because chips are getting bigger and hotter. By Nomura’s estimates, final test time for NVIDIA silicon runs about 4x Hopper’s level for Blackwell and roughly 7x for Rubin. System-level test and burn-in stretch with every generation as well, lifting test’s share of AI chip cost from an estimated 1.9% at Hopper to 3.3% at Rubin.
The mechanics are simple. More transistors mean more items to check; higher power means testing under active thermal management. To ship the same number of chips, you need more testers.
Two companies effectively keep this gate: Advantest and Teradyne ($TER). In June I kept only the former as a seed. The latter was already trading ahead of where I thought the business stood, so I left it on the watch list. Teradyne has since resurfaced in my Physical AI deep dive as the parent of Universal Robots, which gives me one more reason to line the two up side by side.
Advantest: the answer came the morning after
Results landed after the close on July 29. Revenue of ¥367.5bn, up 39.3% year over year; operating profit of ¥190.0bn, up 53.3%; an operating margin of 51.7%, a quarterly record across the board. The company raised full-year guidance again, to ¥1,714bn in revenue and ¥846bn in operating profit, explaining that AI inference test demand had already outrun its April assumptions.
Going into the print, the stock was depressed. Contagion from a sharp memory-sector selloff had knocked it down more than 10% the day before, and it closed the release day slightly lower at ¥25,200. On the call, management flagged that Q1 margins had been flattered by one-off items, including an inventory valuation reversal, and warned of memory component cost pressure in the second half.
The market’s answer came the next morning anyway: two consecutive double-digit up days, to ¥32,500 as of July 31.
Notice the speed gap. The stock is up 19% since June, but forward earnings estimates climbed faster still. The target moved along with the price, and each dollar of earnings costs less today than it did in June. From the perspective of the seed planted in June, the verification event confirmed the thesis, and the price wasted no time reflecting it.
Two triggers to watch from here.
The first is the Q2 operating margin at the late-October print. My margin assumption sits below guidance, so this release is less an upside catalyst than a referee. If margins hold through the cost pressure management itself warned about, the re-rating stands. If they break badly, part of it unwinds.
The second is where analyst targets settle. I am watching whether the average lingers near my coordinates — where a price sits against what I think the business is worth — or starts running toward the aggressive end of the range.
Teradyne: where the market got there first
A day earlier, on July 28, Teradyne also delivered a strong quarter. Revenue of $1,329M, up 104%, and adjusted EPS of $2.47 beat consensus. More decisive was the Q3 revenue guide: $1.25B at the midpoint, far above the $1.03B consensus. The company itself put the second-half fade scenario to rest. The stock rose 6.2% on the first trading day after the release and has since climbed to $365.
Teradyne also owns Universal Robots, which keeps it on my Physical AI watch list. Robotics revenue reached $100M this quarter, up 33% year over year and a fifth consecutive quarter of growth, with electronics manufacturing and semiconductors becoming the group’s largest end market. Automation demand from the factories building out AI data centers is starting to reach the robot arm. At 8% of total revenue, though, it is not yet a variable that moves the coordinates.
The coordinates here tell the opposite story to Advantest’s. In June the price had run ahead of the business. Since then the price came down and earnings came up, closing the gap from the other side. This is a case of fundamentals catching up while the price waited.
Two triggers here as well.
The first is whether robotics keeps growing. Management said the segment should grow in line with the company over the midterm, a claim that gets tested every quarter. The humanoid-driven Physical AI option has yet to appear in either the numbers or the language of the call; for now it lives outside the numbers.
The second is the relative multiples of the two companies. Over the past six weeks the spread blew wide open and then closed almost entirely this week. If it opens again, I will want to know whose circumstances changed.
Two companies at the same gate
One sentence each. Advantest: the seed sprouted, and the price caught up with the thesis in two days. Teradyne: the price had long since arrived, and this week the fundamentals caught up with the price.
So which season is this? A seed spot requires two conditions: the market must not yet see it, and the distance must still be open. Both names now sit outside those conditions. Prices have reached the neighborhood of my conservative case, and the field is already crowded with watching eyes.
Nor is it harvest season. Both names still have room before they reach my aggressive case. Between planting and harvest lies the growing season, and in the growing season my job is neither to plant more nor to rush the harvest, but to water the field and watch the gauges.
There are two gauges.
One: the market’s temperature. Let me be honest about something first. These are names I entered late; analyst targets were already thick on the ground when I arrived. Unlike fields where I set my coordinates first and watched the market approach, I cannot claim here that consensus followed me, and I won’t. What I can do is use my band as a thermometer, and ask where the market’s eyes sit between my conservative scenario and my aggressive one.
The reading today: on Advantest, the analyst average sits barely above my conservative case, and even the most bullish target stays inside my aggressive one. The market’s eyes are gathered in the lower half of the band. Moderate. On Teradyne, the average runs well into the upper half, and the most bullish target clears my aggressive case entirely. One notch warmer. In my experience, the moment a reading like that starts sprinting toward the top of the band usually marks the early edge of overheating.
Two: Advantest’s margin proof in late October. This is the one place in this field where I differ from the market. My margin assumption is more conservative than guidance, my coordinates are built on that conservative assumption, and the current price sits below even that. Which makes this checkpoint asymmetric for me. If margins hold as guided, my band moves up. If my caution proves right, the current coordinates don’t move. The one scenario that demands care is cost pressure breaking below even my conservative floor, in which case I will lower the band and show you the outcome, exactly as it lands, in the next tracker.
To put my cards on the table: both companies came out of this quarter with a firmer hold on the gate. But after the sharp post-earnings run, the story you would start today, at this price, is not the one that was available before the prints. Even so, I read this move as the market recognizing what test is worth and the price catching up to that recognition. In Kostolany’s language, I suspect this sector is still held by the steadfast hands rather than the trembling ones. When the temperature climbs and the composition of hands begins to change, that is when I will be checking these gauges far more often.
The June thesis, that test is a core bottleneck of this cycle, was confirmed this week by earnings and by price. A field that has sprouted stays under watch. The next check is October.
Everything above covers why the gate narrows and where each company stands in front of it. If you want to know exactly where the needles on those gauges are pointing, the coordinates and the bands behind them are in the full research.
Ciao. 🌱
This newsletter is for informational purposes only and does not constitute investment advice or a recommendation to buy or sell any security. The author may hold positions in the securities discussed. Do your own research and consider your own circumstances before investing.
🌱 Growth Wave · Full Deep Dive
[Tracker] Advantest · Teradyne | The Gate Every Chip Must Pass
The article above is a summary. Subscribers get the full valuation bands, the coordinates behind both gauges, and the October checkpoint in detail.