Newsletter · · Ashutosh Agarwal
Three Chip Equipment Deals Put Advanced Packaging Back in Focus - The Semiconductor Equipment Brief - Week of May 31, 2026
Wafer-fab-equipment newsletter for the week of May 31, 2026. Three deals land at once (AMAT carving out ASMPT's panel-level deposition, Onto taking a stake in Rigaku, and the Excelis-Veeco merger clearing final hurdles) pulling advanced packaging back into focus on an otherwise one-sided tape.
The Semiconductor Equipment Brief
Week of May 31, 2026: Three Chip Equipment Deals Put Advanced Packaging Back in Focus
Three deals in one week dragged advanced packaging back into the spotlight: Applied Materials carving out ASMPT's panel-level deposition business, Onto Innovation taking a quarter of Rigaku's X-ray metrology franchise, and the long-pending Excelis-Veeco merger finally pushing through final regulatory hurdles. None of these are blockbusters in isolation. Together, they sketch out where the WFE incumbents see the next leg of fab spending: not on the lithography stage where ASML already monopolizes the conversation, but on the packaging side, where AI workloads are forcing toolmakers to treat heterogeneous integration as a real revenue category. Operators stayed off the mic this week, so the read here is necessarily a pundit framing, and we flag that up front.
TL;DR
- AMAT is bolting on ASMPT's "NEXT" panel-level packaging deposition franchise, a deliberately narrow carve-out to dodge antitrust scrutiny while extending the AMAT advanced-packaging stack.
- Onto Innovation took a 27% stake in Rigaku (X-ray metrology for 3D-stack and advanced-packaging QC), funded by convertible debt: quiet consolidation in the most defensive corner of WFE.
- The Excelis-Veeco merger is in final approval, awaiting China sign-off, with combined revenue at close estimated at ~US$1.5B vs. ~US$1.7B combined 2024, confirming a cycle trough through the merger window.
What's New
Packaging gets the M&A action, not litho. Nicholas Rossolillo on the Chip Stock Investor Podcast (pundit/host) walked through Applied Materials' announced acquisition of ASMPT's "NEXT" segment, the deposition franchise focused on panel-level (square, not round) substrates. The framing was deliberate: Rossolillo argued AMAT structured this as a narrow carve-out specifically because buying ASMPT outright "would definitely run into antitrust issues." It's a way to consolidate share in panel-level packaging tools (the bottleneck behind every AI accelerator that needs 3D stacking) without inviting a regulator review.
Onto-Rigaku is the quieter, more interesting trade. Same episode: Onto Innovation took a 27% equity stake in Rigaku, the Japanese X-ray imaging house whose tools sit at the center of 3D-stack and advanced-packaging QC. Convertible-debt funded. Rossolillo's view was that metrology remains "one of the more durable segments" of WFE (recurring service revenue, software attach) and that ONTO "is still an interesting one to us here" despite elevated multiples. The Rigaku move isn't a takeout; it's a foothold in an adjacent metrology specialty Onto can later consolidate or partner around. Worth flagging given Onto's earlier Kulicke & Soffa equipment carve-out: the company is methodically building a packaging-metrology suite.
Excelis + Veeco crossing the line at the trough. Rossolillo flagged that the Excelis-Veeco merger is in final-approval stages, with China regulatory sign-off the last gate and an expected H2 2026 close. The number worth holding in your head: combined revenue at close estimated at ~US$1.5B, versus a 2024 combined figure of US$1.7B. That's a deal closing into a cyclical revenue trough, the kind of bottom-of-cycle consolidation that historically produces the cleanest synergy math on the other side, if you believe the broader WFE up-cycle thesis. Pundit framing only: no operator voices on the merger this week.
The Debate
This week's tape, frankly, only voiced one side. Rossolillo's stance is constructive: "26 and 2027 are going to be record revenue years" for the Fab Five (ASML, AMAT, LRCX, TEL, KLAC), which he claims still command "about 70% of annual fab equipment spending" (unverified pundit figure). He flagged "a hiccup around 2028… not necessarily going to be like a bear market… but maybe some sort of mid-bull market cycle downturn for semiconductors," followed by a resumed leg higher into end of decade. The demand-side underpinning was a paraphrase of recent TSMC commentary: "we're going to need a lot more semiconductors for the next three to five years. Please keep expanding fab capacity."
The bear case (Intel 18A actually catching, Samsung GAA closing the yield gap, a China pull-forward unwinding, mature-node digestion, Taiwan-Strait tail risk) did not get voiced on the pods this week. That doesn't make it wrong; it means no guest argued it, and we won't manufacture a counter-source to make the section feel balanced. Read the bull framing accordingly: it's pundit consensus filling a quiet tape, not stress-tested by an adversarial operator.
Read-throughs
- CoWoS / advanced packaging. AMAT-ASMPT NEXT and Onto-Rigaku both signal that incumbent WFE players see panel-level packaging and 3D-stack metrology as the next durable revenue category. Watch how AMAT presents NEXT in its first post-close quarter: segment disclosure here will tell you whether they're chasing a meaningful TAM or buying optionality.
- Metrology defensibility. Onto's methodical roll-up (Kulicke & Soffa equipment, now Rigaku) reinforces the thesis that metrology is the most recurring, services-rich pocket of WFE. Pair-trade implication for those who want it: long ONTO / short pure-deposition-and-etch names if you want AI-packaging exposure without the cyclical bookings volatility.
- No operator color on EUV, High-NA, 2026 WFE dollar prints, leading-edge intensity, or China mix. ASML, LRCX, and KLAC appeared this week only inside aggregate Fab-Five framing. Bigger reads on these threads will likely surface around the next earnings cycle.
- Quantum hardware: silent. No IBM Starling, Google Willow, IonQ, Rigetti, PsiQuantum, Quantinuum, D-Wave, or Atom Computing clips on the week's tape. The long-dated frontier thread (quantum roadmaps pulling on advanced packaging and cryogenic control) went uncovered.
What Changed
The marginal new fact this week is the cluster of WFE M&A: AMAT-ASMPT-NEXT, Onto-Rigaku, and Excelis-Veeco's final-approval phase all surfacing together. M&A clustering at a cycle trough is a more reliable signal than any single deal in isolation; it usually means buyers see the bottom and are getting in front of the next leg. We'd weight that more heavily than the pundit "record 2026/27" framing it traveled alongside.