[LRCX] Lam Research Thesis 2026: HBM + 3D NAND Vertical Etch Leadership Tests Memory Cycle Recovery
Lam Research FY2025 (Jun 2025) revenue ~$15.7B (+10-12% YoY) recovering from FY2024 memory capex trough ($14.9B). Adj. EPS ~$3.55 post 10-for-1 stock split (Oct 2024). HBM-related equipment revenue grew ~6-7x from $0.5B FY2022 to $3.0-3.5B FY2025 as HBM bit production share expanded 5-8% → 30-35% of total DRAM bits. Lam Research positioned as key supplier to SK Hynix + Samsung + Micron HBM production with HBM-specific TSV etch (~60-70% market share). 3D NAND vertical channel etch ~70%+ market share — strategically protected by multi-decade R&D investment competitors haven't replicated. Advanced logic 32% of mix (up from 22% FY2022) on AI cycle. FY2026 thesis: HBM4 ramp + HBM3E sustained drives continued DRAM equipment growth; 3D NAND industry recovery from FY2024 trough; advanced logic GAA equipment demand; key risks: China export controls (30-35% revenue), AI cycle reset, memory overcapacity reversal.
Key Takeaways
Lam Research Corporation's fiscal year 2025 (52-week year ended June 29, 2025; Lam Research uses a fiscal calendar that closes near the conclusion of the semiconductor industry's first half capital expenditure cycle) was the year that demonstrated the operational leverage of the focused etch and deposition equipment specialist through the convergence of three structural demand drivers: high-bandwidth memory (HBM) capacity expansion at SK Hynix, Samsung, and Micron, 3D NAND flash memory technology transitions, and advanced logic gate-all-around process node introductions: revenue of approximately $15.5-16B (+~10-12% YoY), adjusted operating margin of approximately 28-30% generating operating income of approximately $4.5-4.8B, and adjusted EPS of approximately $3.50-3.75 on approximately 1.27B diluted shares (post 10-for-1 stock split executed October 2024 to improve retail accessibility). The strategic identity that distinguishes Lam Research from peer semiconductor equipment companies (Applied Materials covered separately as platform-breadth leader, ASML in lithography, KLA in process control, Tokyo Electron) is the deliberate concentration in etch and deposition — Lam Research focuses substantially all R&D and product development on these two process categories where the company has built market leadership through multi-decade technology investment. The investment thesis for Lam Research in FY2026 (fiscal year ending late June 2026) centers on three structural questions: (1) whether the HBM memory capacity buildout — Lam Research is positioned as a key supplier to SK Hynix, Samsung, and Micron HBM production lines, with HBM-specific etch and deposition equipment carrying premium economics — sustains through the FY2026-FY2028 period as HBM3E ramp completes and HBM4 begins; (2) whether the 3D NAND technology transitions (industry progressing toward 300+ layer 3D NAND structures requiring advanced etch capabilities for high-aspect-ratio channel hole formation, where Lam Research holds dominant market share) drive sustained NAND equipment demand; and (3) whether the advanced logic capital expenditure cycle drives Lam Research etch and deposition demand at TSMC, Samsung, and Intel as gate-all-around transistor architectures require new etch/deposition processes that disproportionately benefit specialist suppliers.
Lam Research was founded in 1980 in Fremont, California by David Lam to commercialize plasma etch technology for semiconductor wafer processing. The 45-year operational history transformed Lam Research from a niche etch equipment company into the dominant global etch equipment company plus a major deposition equipment provider, with strategic milestones including the 1997 acquisition of OnTrak Systems (CMP — subsequently divested in 2004 as Lam refocused on etch and deposition), the 2007 acquisition of Bullen Industries (single-wafer cleaning equipment), the 2012 acquisition of Novellus Systems ($3.3B — the transformative deal that added comprehensive deposition capability to Lam's etch leadership), the 2016 attempted merger with KLA-Tencor (the deal was abandoned in 2016 after multi-jurisdictional antitrust scrutiny), and the strategic positioning as the focused etch + deposition specialist. CEO Tim Archer, who has led Lam Research since December 2018 (succeeding Martin Anstice), oversaw the strategic positioning that emphasized memory market leadership (3D NAND vertical channel etch, DRAM specialty etch, HBM-specific equipment), advanced logic gate-all-around process technology positioning, and the operational discipline that has supported industry-leading R&D investment (approximately $1.7B annually, approximately 11% of revenue) at the same time as consistent shareholder returns. The strategic identity that distinguishes contemporary Lam Research from peer semiconductor equipment vendors is the deep technology focus — Lam concentrates 11% of revenue on R&D specifically in etch and deposition technologies, building competitive advantages through cumulative engineering investment that platform-breadth competitors cannot replicate at this depth.
Business Structure
Lam Research operates as a focused etch and deposition equipment company without formal segment reporting (the company reports a single consolidated business segment, with revenue disclosed by product category and end-customer mix in the management commentary).
Etch (~$8-9B revenue, ~55-60% of total): Lam Research's core technology heritage and primary market leadership position. Etch product categories:
- Reactive Ion Etch (RIE): The traditional plasma etch process used across most semiconductor process steps — Lam holds approximately 50%+ market share globally.
- Atomic Layer Etch (ALE): The advanced atomic-precision etch process critical for advanced logic and 3D NAND applications — Lam holds approximately 60-70% market share, supported by multi-decade technology investment.
- Specialty Etch Applications: 3D NAND vertical channel etch (the high-aspect-ratio etch process critical for stacking memory cells in 3D NAND structures, where Lam Research holds approximately 70%+ market share — the most strategically distinct etch technology in the company's portfolio), DRAM specialty etch, advanced packaging etch (HBM through-silicon vias, wafer-level packaging), MEMS/sensor etch.
The etch equipment opportunity scales with two technology vectors: (1) wafer process step count expansion as semiconductor node transitions add more etch steps (each generation typically adds 10-20% more etch steps), and (2) the 3D vertical scaling in memory technology that requires substantially more etch-intensive processing.
Deposition (~$5-6B revenue, ~35-40% of total): Inherited from the 2012 Novellus acquisition plus subsequent organic development. Deposition product categories:
- Chemical Vapor Deposition (CVD): Traditional CVD plus emerging plasma-enhanced CVD (PECVD) for various dielectric and metal film deposition applications.
- Atomic Layer Deposition (ALD): Lam Research has been investing significantly in ALD technology for advanced logic gate-all-around process technology applications and selected memory applications.
- Electroplating: Copper electroplating for chip interconnect and emerging applications including advanced packaging.
- Selective Deposition: Emerging selective deposition processes that enable bottom-up build-up of specific materials with patterning advantages for advanced nodes.
Customer Support Business Group (CSBG) (~$2-3B revenue, ~15-20% of total): Service contracts on installed Lam equipment, spare parts, equipment refurbishment, training services, plus selected used equipment sales. CSBG operates with favorable margins and grows steadily as the installed base expands.
Key Core Metrics Performance
Revenue, Margin, and EPS Trajectory (FY2021–FY2025, fiscal years ending late June)
| Fiscal Year | Revenue | Adj. Op. Margin | Adj. EPS (post-split FY2025) | Diluted Shares (post-split) |
|---|---|---|---|---|
| FY2021 (Jun 2021) | ~$14.6B | ~31.5% | ~$3.27 | ~1.45B |
| FY2022 (Jun 2022) | ~$17.2B | ~30.5% | ~$3.65 | ~1.40B |
| FY2023 (Jun 2023) | ~$17.4B | ~30.0% | ~$3.34 | ~1.34B |
| FY2024 (Jun 2024) | ~$14.9B | ~28.5% | ~$2.86 | ~1.30B |
| FY2025 (Jun 2025) | ~$15.7B | ~29.5% | ~$3.55 | ~1.27B |
Note: Lam Research executed a 10-for-1 stock split effective October 2024; all per-share data shown post-split for comparability. Pre-split FY2024 EPS would have been approximately $28.55 versus post-split $2.86; FY2025 post-split $3.55 equivalent to pre-split $35.50.
The pattern of revenue compression in FY2024 (approximately -14% from FY2023 peak of $17.4B to $14.9B) reflects the memory industry capital expenditure trough as DRAM and NAND manufacturers cut capacity additions to support memory pricing recovery. The FY2025 modest recovery to approximately $15.7B reflects the early stages of memory capex recovery plus advanced logic equipment demand from the AI infrastructure cycle.
End-Market Mix Evolution
| End Market | FY2022 % of Revenue | FY2024 % of Revenue | FY2025 % of Revenue |
|---|---|---|---|
| DRAM (incl. HBM) | ~25% | ~28% | ~30%+ |
| 3D NAND | ~32% | ~22% | ~20% |
| Advanced Logic | ~22% | ~28% | ~32% |
| Mature Logic | ~10% | ~12% | ~10% |
| Foundry (multi-node mix) | ~11% | ~10% | ~8% |
The mix shift toward advanced logic and DRAM/HBM (combined approximately 62% of revenue in FY2025 vs approximately 47% in FY2022) reflects the AI infrastructure cycle dynamics: advanced logic capex scaling at TSMC + Samsung + Intel for AI accelerator manufacturing, plus DRAM/HBM capacity expansion at SK Hynix + Samsung + Micron supporting AI accelerator memory requirements. 3D NAND mix has declined as NAND industry consolidation plus pricing-driven capacity restraint compressed NAND equipment demand.
HBM Memory Equipment Position
| Period | HBM Bit Production (% of Total DRAM Bits) | Lam HBM-Related Revenue | Lam Memory Customer Mix |
|---|---|---|---|
| FY2022 | ~5-8% | ~$0.5B | DRAM ~25%, NAND ~32% |
| FY2023 | ~10-15% | ~$1.0-1.5B | DRAM ~28%, NAND ~28% |
| FY2024 | ~20-25% | ~$2.0-2.5B | DRAM ~28%, NAND ~22% |
| FY2025 | ~30-35% | ~$3.0-3.5B | DRAM ~30%+, NAND ~20% |
The HBM bit production share growing from approximately 5-8% in FY2022 to approximately 30-35% in FY2025 reflects the rapid HBM capacity expansion. Lam Research's HBM-related equipment revenue has grown approximately 6-7x over this period, supporting elevated DRAM segment economics for Lam Research and the consolidated business.
Market Evaluation
Lam Research trades at approximately 22-28x forward adjusted EPS — premium semiconductor equipment multiples that reflect both the technology focus quality and the AI infrastructure cycle exposure. The bull case is HBM continued growth + advanced logic gate-all-around demand + 3D NAND recovery: if HBM bit production share continues expanding toward 40-50% of total DRAM bits by FY2027 (HBM4 ramp + HBM3E continued production), if advanced logic capital expenditure cycle continues at TSMC + Samsung + Intel, and if 3D NAND industry recovers from FY2024 trough as NAND pricing normalizes, total revenue could approach $20-22B with adj. EPS approaching $4.80-5.40 by FY2027 — supporting equity at sustained 22-26x. The bear case is China export control extension + AI cycle reset + memory overcapacity: if China export controls extend further restricting Lam Research's trailing-edge equipment sales (China revenue approximately 30-35% of Lam total), if AI infrastructure demand resets compressing advanced logic capacity expansion, or if memory industry cycles into overcapacity again with both DRAM and NAND equipment demand compression, revenue could decelerate with EPS compression.
The HBM Memory Position and 3D NAND Etch Leadership
The strategic argument that frames Lam Research's contemporary investment thesis rests on two distinct memory technology positions: HBM (high-bandwidth memory) capacity buildout and 3D NAND vertical channel etch leadership. Both positions reflect Lam Research's deep technology investment and the structural premium economics that specialty etch and deposition equipment commands.
The HBM position: HBM is the AI accelerator-critical memory technology used in NVIDIA H100/H200/B100/B200, AMD MI300, Google TPU, and emerging custom AI silicon. HBM is manufactured by stacking 4-12 DRAM dies vertically with through-silicon vias for inter-die communication, requiring specialized process equipment for: high-aspect-ratio TSV etch (where Lam Research has approximately 60-70% market share), wafer-level packaging deposition, advanced bonding processes, plus selected emerging HBM-specific process steps. The HBM bit production share growing from approximately 5-8% in FY2022 to approximately 30-35% in FY2025 (and projected toward 40-50% by FY2027 as HBM4 ramps) supports continued elevated demand for Lam Research's HBM-specialized equipment.
The 3D NAND position: 3D NAND flash memory progressively scales by adding vertical layers of memory cells (industry currently producing 200+ layer products with roadmaps targeting 300+ layers by FY2027). The vertical channel etch — drilling high-aspect-ratio holes through the entire 3D NAND structure to enable memory cell formation — is the most technically demanding 3D NAND process step. Lam Research holds approximately 70%+ market share in 3D NAND vertical channel etch, supported by multi-decade technology investment in deep silicon etch capabilities that competing equipment vendors have not been able to replicate at the same depth. As 3D NAND scales beyond 300 layers, the etch technology challenges progressively intensify, supporting Lam Research's continued market share dominance in this critical process step.
The competitive dynamic: Applied Materials competes with Lam Research in etch (Applied Materials etch revenue approximately $5-6B versus Lam etch revenue approximately $8-9B), plus Tokyo Electron in selected etch applications. The specialty etch segments where Lam Research holds dominant market share (3D NAND vertical channel etch, atomic layer etch, advanced logic high-aspect-ratio etch) are strategically protected by the cumulative R&D investment Lam has deployed over multiple decades — competitive displacement at these specialty applications would require multi-year competitive R&D investment that competing equipment vendors have not been demonstrating capability to execute at scale.
