ACM Research (Shanghai), Inc. (688082.SS): The Megasonic Pioneer Clean-Rooming China's Semiconductor Ambitions

I. Introduction & Episode Roadmap

On December 2, 2024, the United States Bureau of Industry and Security added a company to its Entity List. The designation meant that no party anywhere in the world could supply that company with hardware, software, or technology subject to US export jurisdiction without a license Washington had no intention of granting.

The company was 盛美半导体设备（上海）股份有限公司 ACM Research (Shanghai), Inc., together with its operating subsidiaries in mainland China and Korea. Its majority owner — ACM Research, Inc. of Fremont, California — was not listed. Neither were its subsidiaries outside the mainland. The parent later stated that ACM Shanghai and ACM Korea had not been notified of any specific wrongdoing that led to the designation.¹

Consider that corporate structure again: a California company founded by a Silicon Valley engineer owns roughly three-quarters of a Shanghai-listed subsidiary that the American government has blacklisted. The parent files with the SEC; the subsidiary files with the 上海证券交易所 Shanghai Stock Exchange. Substantially all of the revenue comes from customers in mainland China.² Meanwhile, the two listed securities representing claims on largely the same operating business trade at valuations separated not by percentage points, but by multiples.

What the company actually does. Manufacturing a modern semiconductor requires depositing, patterning, and etching dozens of layers onto a silicon wafer. Between almost every step, the wafer must be cleaned of particles, metal residues, and leftover photoresist. Contamination invisible under an optical microscope can ruin an entire die. As feature sizes shrank below 45 nanometers, wafer cleaning evolved from a routine utility into one of the highest-frequency, highest-stakes steps in a fabrication plant, repeated thirty to forty times per wafer.

However, a wafer fab's two main cleaning requirements conflict directly. Cleaning aggressively risks damaging the delicate three-dimensional structures built over preceding weeks. Cleaning gently leaves microscopic particles behind. ACM's core technical insight was a method to reconcile this trade-off using controlled acoustic energy.

The numbers, as of today. For 2025, ACM Shanghai reported revenue of ¥6.786 billion, up 20.8%, with a gross margin of 48.32% and net profit attributable to shareholders of ¥1.396 billion, up 21.05%.³ The parent company, reporting in dollars under US GAAP, recorded $901.4 million of revenue for the same year.² By late August 2026, the Shanghai-listed shares carried a market capitalization of roughly ¥143 billion, while the Nasdaq-listed parent's entire market capitalization stood at about $4.8 billion.⁴⁵

The core thesis, and the tension inside it. ACM's disclosed share of the Chinese single-wafer cleaning equipment market reached 23% in 2024, and Gartner data cited by the company placed its 2023 global cleaning market share at 6.6% — fifth worldwide in a segment where four foreign incumbents controlled over 90%.⁶ The central question is whether that market position, established through process technology and accelerated by domestic supply-chain localization, can be expanded into vertical furnaces, coater-developer tracks, electrochemical plating, and thin-film deposition — equipment categories where ACM lacks an established base, faces qualified incumbents, and competes against well-capitalized domestic equipment makers.

A recent quarterly trend highlights this transition challenge. In the June 2026 quarter, ACM's consolidated cleaning revenue — its core product line — fell 14.2% year over year, while non-cleaning product categories grew at triple-digit rates.⁷ This raises a critical operational question: does the shift reflect a planned expansion into broader platform tools, or an accelerating competitive displacement in its core cleaning market before adjacent product lines reach commercial scale?

The roadmap. The analysis begins with a Silicon Valley startup that spent nearly a decade commercializing its initial technology before securing market adoption. It follows the strategic shift to Shanghai and the development of the acoustic cleaning methods that established the company's market footprint. It examines the multi-year qualification process at a tier-one Korean memory manufacturer, detailing what that commercial win validated. It analyzes the STAR Market listing, the dual-listed capital structure, and the short-seller report of 2020. It evaluates the geopolitical environment in both directions — as an initial catalyst for domestic market adoption and as an ongoing constraint on global supply chain operations. It assesses the platform expansion strategy, distinguishing recognized equipment revenue from evaluation tools deployed at customer fabs. Finally, it outlines the competitive dynamics, the bull and bear investment theses, and the operational metrics that will determine ACM's long-term market position.

The narrative starts with an initial technology that offered technical performance advantages but faced adoption hurdles in the commercial fab market.

Subscribe on LinkedIn

↑ Back to Top

II. Silicon Valley Roots, Copper Dreams, & The Pivot to Wet Cleaning (1998–2008)

In January 1998, 王晖 Dr. David H. Wang founded ACM Research in Fremont, California, focusing on copper interconnect technology.⁸

Wang possessed strong technical credentials. He held a bachelor's degree in precision instruments from 清华大学 Tsinghua University along with a master's and a doctorate in precision engineering from 大阪大学 Osaka University, and he had worked in semiconductor equipment before founding the company.⁸ According to his official biography, Wang is the inventor of stress-free copper polishing and holds more than one hundred patents.⁸

His strategic focus aligned with a major industry transition. Semiconductor manufacturers were switching interconnect wiring from aluminum to copper, which offers superior electrical conductivity but presents processing challenges. After depositing copper across a wafer, fabricators must remove excess metal to create a flat surface. The industry's solution was chemical mechanical planarization, or CMP, which grinds the wafer surface flat using an abrasive chemical slurry pressed against a rotating pad.

While effective, CMP applies significant mechanical force. On wafer stacks containing fragile, low-dielectric-constant insulating layers, that force risks cracking or delaminating thin films. Wang's stress-free polishing substituted electrochemistry for mechanical abrasion, dissolving excess copper using an electric field and an electrolyte bath. The process required no downforce, used no abrasive slurry, and eliminated mechanical stress on dielectric layers.

Why it failed anyway, and what that failure teaches. Despite its technical logic, the process failed to achieve commercial adoption. Semiconductor manufacturers standardized on conventional CMP equipment supplied by established vendors like Applied Materials and Lam Research, leaving ACM's initial copper polishing tool without a meaningful market.⁹ The company survived on investment capital rather than equipment sales and eventually redirected its development efforts toward wafer cleaning.⁹

This initial setback highlights a structural challenge that bears directly on ACM's current platform expansion.

ACM possessed a novel physical approach to a recognized manufacturing bottleneck, backed by proprietary patents and deep founder expertise. However, it lacked a semiconductor fabricator willing to risk a production line on an unproven tool. Semiconductor manufacturing operates as an integrated system rather than a collection of standalone machines: altering a planarization step disrupts downstream metrology, defect signatures, etch steps, and yield models refined over years of production. Evaluating a new process technology requires fab managers to weigh a theoretical performance advantage against the risk of disrupting a proven yield curve. Incumbent toolmakers typically win that calculation unless a newcomer's economic or technical advantage is overwhelming.

ACM spent nearly a decade absorbing this operational reality. The experience established a clear precedent in the company's corporate history: superior process physics is a prerequisite, but it does not guarantee commercial adoption. That lesson remains relevant as ACM seeks to expand into track and thin-film deposition markets against entrenched incumbents.

Stress-free copper polishing never achieved material scale. Today, it survives as a minor line item folded into ACM Shanghai's "other semiconductor equipment" disclosures alongside plating and furnace tools.¹⁰ Nearly three decades after its founding, ACM's original core technology remains a negligible portion of overall revenue—illustrating how difficult it is to displace established process standards in semiconductor fabrication.

The move east. By the mid-2000s, the geographical center of semiconductor manufacturing was shifting toward Asia. New fab construction concentrated in the region, engineering talent in Shanghai offered significant cost advantages over Silicon Valley, and Chinese policymakers began treating domestic semiconductor capability as a strategic priority.

In 2005, ACM established ACM Research (Shanghai), Inc., locating its research, engineering, and manufacturing operations at the Zhangjiang High-Tech Park.¹¹ The move represented a fundamental shift in operational focus: the company moved its primary development work, engineering staff, and target customer base to China, while maintaining its parent corporate registration in California.

Finding the real bottleneck. ACM next targeted megasonic wafer cleaning. The underlying mechanism relies on passing high-frequency sound waves through a liquid chemical bath, generating microscopic bubbles that expand and collapse. This acoustic cavitation dislodges microscopic contaminants that physical brushes cannot remove at nanometer scales.

However, uniform acoustic energy distribution proved difficult to achieve. Traditional bath systems created standing waves with alternating hot spots and dead zones. In dead zones, contaminants remained on the wafer; in hot spots, intense bubble collapse released enough localized energy to damage delicate, high-aspect-ratio silicon structures. As device feature sizes shrank, the operating window between insufficient cleaning and structural destruction narrowed significantly.

ACM addressed this limitation by developing Space Alternated Phase Shift, or SAPS. Rather than applying a static acoustic field, SAPS alternates the phase of the megasonic waves while moving the transducer relative to the rotating wafer, ensuring every point on the surface receives an identical acoustic dose over time.⁹ The technique provided uniform energy distribution across the wafer surface without damaging delicate circuit features.

SAPS delivered a measurable technical improvement to a cleaning step repeated dozens of times per wafer. To turn that technology into a commercial business, however, ACM needed a major semiconductor manufacturer to validate the tool in high-volume production—and in 2008, no mainland Chinese fab possessed the operational scale or industry standing to serve as a global reference customer.

↑ Back to Top

III. Qualifying Tier-1 Fabs & The Technical Moat: SAPS, TEBO, & Tahoe (2008–2020)

The commercial breakthrough arrived via Wuxi, China, where Korean memory maker SK하이닉스 SK Hynix had established a 300-millimeter DRAM fabrication facility. The Wuxi fab evaluated ACM's single-wafer cleaning systems and began placing commercial orders around 2010, making SK Hynix ACM's first major customer.⁹ For an equipment startup without a high-volume production track record, qualifying at a tier-one memory fab provided critical industry validation.

The mechanics of tool qualification. In semiconductor capital equipment, winning a qualification requires far more than delivering a functional machine. Fabs evaluate process results rather than standalone hardware.

When ACM ships a tool for evaluation, the equipment remains recorded as ACM's inventory on the customer's fab floor. Engineers from both companies run comparative split lots over several months, evaluating wafer samples processed through the new tool against existing production baselines to track particle counts, pattern integrity, and electrical yield consistency. Only after the equipment demonstrates repeatable performance across product types is it written into the fab's official process of record, allowing ACM to recognize the associated revenue.

This qualification cycle explains key structural dynamics in ACM's financial reporting. Finished goods inventory remains elevated primarily because evaluation tools are deployed at customer sites awaiting final acceptance.⁷ Consequently, quarterly equipment shipments and recognized revenue frequently diverge over multi-quarter periods. Once an initial tool passes qualification, however, follow-on orders for the same process line require minimal additional engineering validation, driving high repeat-order rates.

These qualification barriers also illustrate why entrenched equipment vendors remain difficult to displace. Fab managers prioritize proven yield stability over marginal tool improvements. Introducing a new vendor requires absorbing months of engineering overhead and accepting potential yield disruption, meaning a challenger must offer substantial technical or economic advantages to justify replacing an incumbent process.

Scope and limits of commercial adoption. While the SK Hynix qualification validated ACM's process capability, it did not trigger broad market displacement of incumbent vendors.

SAPS technology did not gain traction in leading-edge FinFET logic fabrication, where pattern collapse risk is highest.⁹ ACM subsequently developed its second-generation megasonic cleaning process, Timely Energized Bubble Oscillation, or TEBO, which secured adoption in NAND flash manufacturing at SK Hynix and 长江存储 YMTC.⁹ Although these represented major commercial wins in specific memory applications, they did not translate into general replacement of established suppliers across broader fab operations.

Global market share data underscores this limitation. Four major incumbents—SCREEN Holdings, Tokyo Electron, Lam Research, and Samsung's captive supplier SEMES—collectively controlled over 90% of the global wafer cleaning equipment market, while ACM held a 6.6% share in 2023 according to Gartner data.⁶ More than fifteen years after deploying its megasonic cleaning technology, ACM's global market presence remained concentrated in single-digit territory.

The commercial trajectory indicates that while SAPS and TEBO provided genuine process differentiation in targeted memory applications, novel acoustic technology alone did not overcome incumbent vendor lock-in across global fabs. ACM's broader expansion would ultimately rely on domestic supply-chain policies alongside its core technical capabilities.

TEBO physics and structural cleaning. Where SAPS controlled the spatial distribution of acoustic energy, TEBO addressed the dynamics of cavitation bubble behavior. In standard megasonic cleaning, micro-bubbles expand until they collapse, generating localized shockwaves. While effective on flat surfaces, these cavitation implosions can damage fragile 3D NAND vertical structures and advanced transistor gates. TEBO regulates acoustic energy to keep bubbles oscillating continuously at a stable size without collapsing, creating a controlled scrubbing action that prevents structural damage on complex three-dimensional device architectures.

The Tahoe platform and chemical cost reduction. ACM introduced its third major cleaning system, the Ultra C Tahoe, in December 2019. The hybrid architecture combined batch immersion tanks with single-wafer megasonic cleaning, reducing sulfuric acid consumption by at least 80% compared to conventional single-wafer sulfuric-peroxide systems.¹²

In standard photoresist removal, single-wafer processing consumes large volumes of fresh chemical mixtures, creating significant chemical purchasing and environmental disposal costs. The Tahoe platform performs primary chemical processing in a shared bath before completing final cleaning on individual single-wafer chambers, maintaining defect control while substantially lowering chemical consumption and wastewater treatment overhead.

By addressing operating expenditure and environmental compliance alongside cleaning performance, the Tahoe platform provided semiconductor manufacturers with an economic incentive to evaluate ACM against incumbent vendors.

ACM expanded the Tahoe platform's capabilities over time. In August 2026, the company announced that Tahoe had been broadened to include wet etching and monitor wafer reclaim processes, consolidating functions previously requiring separate equipment, and reported that multiple leading semiconductor manufacturers had adopted the expanded platform.⁷

Financial profile and product mix. This product expansion drove substantial revenue growth. In 2024, single-wafer cleaning equipment generated ¥4.057 billion, accounting for 72.2% of ACM Shanghai's total revenue of ¥5.618 billion. Other semiconductor equipment—including electrochemical plating, vertical furnaces, and stress-free copper polishing—contributed ¥1.137 billion (20.2%), while advanced packaging wet equipment generated ¥246 million.¹⁰ Total revenue for 2024 grew 44.5% year over year.¹⁰

ACM maintained gross margins near 48% during this period, indicating that its pricing power rested on proprietary process capabilities rather than low-cost discounting. However, sustaining these gross margins depends on maintaining technological differentiation as domestic competitors qualify competing equipment lines.

By 2020, ACM had established a profitable core cleaning franchise, secured a tier-one international reference, and expanded its customer base in China. To fund its next phase of growth, the company sought access to mainland capital markets, setting up a corporate restructuring to align its California parent structure with Chinese institutional investment.

↑ Back to Top

IV. The STAR Market IPO, Dual-Listing Governance, & The J Capital Short Attack (2020–2022)

On October 8, 2020, J Capital Research published a short-seller report on ACM Research titled "Dirty Business."¹³ Its central allegation was blunt: that the company was over-reporting revenue and profit while siphoning earnings to related parties.¹³

The related party at the center of the dispute was Ninebell, a Korean supplier of robotic handling components for wet process tools. ACM had invested in Ninebell in September 2017, acquiring a 20% equity stake, with David Wang joining its board.¹⁴ J Capital characterized Ninebell as a purported high-technology investee that actually purchased its robotic arms from 安川電機 Yaskawa rather than manufacturing them itself.¹³

ACM's response, filed with the SEC, disputed the allegations point by point.¹⁴ The company stated that the Ninebell stake was acquired to secure a strategic long-term supply partnership, and that Yaskawa had not supplied robotic arms meeting the technical specifications required for ACM's specialized wet process equipment. It also detailed what it described as numerous factual misstatements in the report regarding revenue, tool average selling prices, gross margins, key suppliers, inventory, and other balance sheet line items.¹⁴

How to weigh this, six years later. The most reliable evidence regarding the fraud allegations lies in subsequent operational execution rather than the initial claims. ACM Shanghai's STAR Market listing proceeded as planned. Revenue grew from roughly $157 million in 2020 to $901.4 million in 2025 — a nearly sixfold increase over five years — while the company reported delivering more than 1,435 tools since 2009, including 1,255 repeat orders or accepted systems.¹⁵² A business fabricating financial results at scale rarely withstands six years of concurrent audits across two accounting regimes, a Shanghai Stock Exchange listing review, annual SEC filing cycles, and ongoing scrutiny from state-linked customer fabs that physically receive and install equipment.

While that history provides a reasonable basis for discounting the core fraud claims, it does not erase the structural vulnerabilities highlighted during the episode.

Two ongoing operational features remain relevant. The first is supplier dependency: reliance on specialized Korean and Japanese mechanical components leaves wet-process equipment makers exposed to supply chain bottlenecks. Management confirmed in August 2026 that component lead times had expanded from roughly four to six months, with supply constraints concentrated in Japan and Korea.⁷ The second is governance optics: holding an equity stake in a key vendor whose board includes ACM's founder creates an arrangement that, while legal and fully disclosed, remains inherently more complex to defend than an arm's-length supplier relationship.

Regarding verifiable facts: the public record confirms that the listing proceeded and that ACM issued a formal rebuttal. However, claims that an independent audit affirmatively cleared Ninebell of wrongdoing cannot be substantiated from disclosed filings and should not be assumed.

The listing itself. ACM Shanghai's shares began trading on the STAR Market — the 科创板 board of the Shanghai Stock Exchange — on November 18, 2021, under code 688082.¹⁶ The company issued 43,355,753 shares at ¥85.00 per share, raising gross proceeds of ¥3.685 billion (¥3,685,239,005), with 海通证券 Haitong Securities serving as sponsor and lead underwriter alongside 中金公司 CICC as joint lead.[^17]

The strategic rationale extended beyond capital raising to corporate identity. For a Chinese semiconductor fab evaluating multi-year equipment commitments during a domestic localization initiative, dealing with an entity structured unambiguously as a onshore listed company — governed by local disclosure rules, employing domestic engineers, and offering local equity incentives — provides operational alignment. The STAR Market listing established ACM Shanghai as a domestically listed equipment vendor while preserving parent control.

Monetizing majority control. The parent company's ownership stake has declined through planned equity transactions rather than a single divestment. ACM Shanghai subsequently raised capital through a private placement to specific institutional investors at ¥116.11 per share under a program authorized for up to ¥4.5 billion, diluting the parent's position.¹⁷ On February 6, 2026, the parent transferred 4,801,648 shares — representing a 1.00% equity stake — via an inquiry-based institutional transfer at ¥160 per share, reducing the combined ownership of the parent and its concert parties from 74.84% to 73.7%.¹⁸ That transaction drew 38 institutional bidders and was 1.17 times oversubscribed.¹⁸ ACM Research's 2025 Form 10-K reported a 74.6% ownership interest in ACM Shanghai.²

Consequently, historical references to an 81.5% parent ownership stake are outdated. While parent equity ownership has trended downward, management signaled a near-term pause: asked in August 2026 whether further expansion would be funded by selling Shanghai shares, CFO Mark McKechnie stated that the company had no immediate plans to reduce its position further, having raised $150 million in a registered direct offering in May 2026 that expanded net cash on its US balance sheet to roughly $300 million.⁷

Evaluating the valuation gap. In late August 2026, ACM Shanghai carried a market capitalization of approximately ¥142.9 billion.⁴ A 73.7% equity interest in the subsidiary represents over ¥105 billion in value. By comparison, parent entity ACM Research, Inc. — which holds that ownership stake alongside US operations and its separate net cash position — maintained a market capitalization of approximately $4.8 billion.⁵

At prevailing exchange rates, the US-listed parent trades at a substantial discount relative to the market value of its primary operating asset. Several market factors account for this divergence: the US Entity List designation applies specifically to the operating subsidiary, parent shareholders absorb China regulatory and audit-access risks that domestic onshore investors price differently, and STAR Market semiconductor valuations remain elevated. ACM Shanghai's shares traded between ¥138.01 and ¥459.99 over the trailing twelve months before closing at ¥296.16 on August 28, 2026 — illustrating significant domestic market volatility.⁴

For investors analyzing 688082, two distinct structural factors apply: ACM Shanghai operates as the primary entity holding the company's core engineering teams, patents, and customer relationships, yet it remains the specific corporate entity subject to US Entity List restrictions, trading at onshore valuation multiples that drive the dual-listing pricing gap.

↑ Back to Top

V. Geopolitics & The China Semiconductor Indigenization Wave (2022–Present)

In October 2022, the US Commerce Department published export control rules that redrew the semiconductor manufacturing map. Advanced logic below specified nodes, 3D NAND above defined layer counts, and DRAM below defined half-pitches became effectively off-limits for Chinese production using American equipment, technology, or personnel. Successive regulatory tightenings followed.

For Chinese fabs, the strategic signal was unambiguous: any tool unavailable domestically represented a potential future supply chain vulnerability. Equipment procurement shifted from a purely technical evaluation to a continuity-of-supply imperative. Domestic toolmakers that had struggled for evaluation slots suddenly encountered customers actively seeking qualified local vendors.

This demand trajectory is directly reflected in ACM's financial performance. Consolidated revenue expanded from $388.8 million in 2022 to $557.7 million in 2023, $782.1 million in 2024, and $901.4 million in 2025.¹⁵² At the Shanghai subsidiary level, revenue reached ¥6.786 billion in 2025.³ In August 2026, management raised the midpoint of full-year revenue guidance to a range of $1.125 billion to $1.175 billion—implying 25% to 30% growth—and reported that first-half orders grew 105% year over year.⁷

How demand reached ACM. The commercial transmission mechanism extended beyond strategic procurement policy.

Chinese fabs responded to trade controls through a dual strategy: pulling forward purchases of restricted foreign tools while available, and accelerating evaluation programs for domestic alternatives across every viable process step. Wet cleaning emerged as an immediate target for substitution—it is a high-frequency manufacturing step, does not require the most advanced foreign technology, and ACM already possessed qualified production tools operating in domestic fabs.

Consequently, evaluation slots that previously required years to secure opened rapidly, altering the fab customer's decision matrix. Rather than requiring justification to test a domestic tool, process engineers faced pressure to justify not evaluating one. This structural reduction in evaluation barriers provided a crucial commercial catalyst for challenger vendors.

Escalating regulatory restrictions. On December 2, 2024, BIS added ACM Shanghai along with its mainland China and South Korea operating subsidiaries to the Entity List.¹ Management publicly stated that the impact on its supply chain and mainland manufacturing capacity would remain manageable through alternative supplier networks, while noting that the net effect on domestic revenue would depend partly on customer capital spending plans.¹

While revenue and order momentum continued following the designation, the regulatory environment imposes clear operational constraints. ACM's equipment incorporates American-origin technology, and its engineering personnel operate across multiple international jurisdictions. The company maintains manufacturing operations in South Korea—where tools bound for US customers are currently assembled—and is constructing a demonstration center in Oregon scheduled to open later in 2026.⁷ Each of these operating units functions within export control boundaries subject to administrative revision.

This creates a distinct strategic tension: the trade policies that accelerated ACM's domestic market adoption simultaneously restrict its international expansion. Management has outlined a long-term revenue target of $4 billion, split between $2.5 billion from mainland China and $1.5 billion from international markets.⁷ Achieving the international target requires penetrating semiconductor fabs in Taiwan, South Korea, Singapore, Europe, and the United States while operating an Entity-Listed subsidiary structure.

Customer concentration and market structure. ACM's primary customer base comprises the leading fabricators driving China's domestic supply-chain initiatives, including 中芯国际 SMIC, 长江存储 YMTC, 长鑫存储 CXMT, and 华虹 Hua Hong. By customer segment, 2025 revenue split roughly 59% foundry, logic, and other applications, 27% memory, and 14% advanced packaging and wafer processing.²

Customer concentration metrics improved significantly over the subsequent twelve months. In the first half of 2026, ACM reported only one customer representing more than 10% of revenue, at 12.7%. By contrast, during the first half of 2025, three customers collectively accounted for 49.9% of total revenue.⁷ Management cited this shift as evidence of a broadening customer base, a characterization supported by disclosed segment figures.⁷

Despite reduced customer concentration, fabricators maintain substantial buyer power. State-backed fabs follow explicit multi-sourcing mandates to mitigate operational risk, avoiding sole-source reliance on single domestic vendors for critical process steps. Furthermore, competing domestic equipment manufacturers possess significant capital resources and public listings: 北方华创 NAURA Technology Group maintained a market capitalization above ¥500 billion in late August 2026, 拓荆科技 Piotech stood near ¥194 billion, 芯源微 Kingsemi reached approximately ¥74 billion, and 至纯科技 PNC Process Systems held roughly ¥9.6 billion.⁴ Among these peers, NAURA generates larger revenues than ACM Shanghai while offering a broader equipment portfolio.

Core cleaning franchise performance. Evaluating whether ACM maintains a defensible position in Chinese wet cleaning requires examining recent product segment data. In the June 2026 quarter, consolidated cleaning revenue dropped 14.2% year over year to $133.0 million, declining to 45.4% of total product mix.⁷ Furthermore, ACM Shanghai's disclosed 23% share of China's cleaning equipment market reflects a solid market presence rather than category dominance.⁶

Management attributed the quarterly decline to a product line transition rather than share loss to competitors. Historically, ACM generated minimal revenue from single-wafer sulfuric-peroxide mix (SPM) cleaning tools, which management estimates represent roughly one-third of the total addressable cleaning market. The company shipped several SPM evaluation units during the first half of 2026 and expects total SPM shipments to exceed twenty tools by year-end, with revenue recognition deferred until customer qualification is complete.⁷ Under ACM's revenue recognition accounting, evaluation equipment remains in finished goods inventory until final acceptance, creating a structural lag between tool shipments and recognized revenue.

This accounting framework is supported by a finished goods inventory balance of $287.9 million, which management confirmed consists primarily of evaluation tools deployed at customer fabs.⁷ However, commercial adoption of the SPM platform remains to be demonstrated in reported revenue. While the evidence does not indicate market share loss, it confirms that the core cleaning franchise must demonstrate sustained growth through this product transition over subsequent quarterly reporting periods.

↑ Back to Top

VI. Platformization: Sizing ECP, Furnace, Track, & PECVD

In the June 2026 quarter, ACM shipped its 2,000th electroplating chamber, after reaching 500 in 2022 and 1,500 in 2025.⁷

That sequence illustrates an accelerating trajectory: taking nearly a quarter century to install its first 500 chambers, three years to triple that base, and just twelve months to deliver the latest 500. Whatever the ultimate reach of ACM's platform expansion, its electrochemical plating business is no longer a secondary endeavor.

The portfolio, honestly divided. ACM's product line falls into three distinct tiers with vastly different maturity profiles.

The first tier is the proven core: single-wafer cleaning systems built on SAPS and TEBO, the Tahoe hybrid platform, backside and solvent cleaners, scrubbers, wet etchers, and high-temperature SPM equipment. This segment generated $626.0 million, or 69.5% of total revenue in 2025.²

The second tier is proven-and-scaling: electrochemical plating (ECP) for front-end copper interconnects and advanced packaging, alongside vertical furnaces and adjacent tools. This category contributed $199.6 million, or 22.1% of 2025 revenue, while advanced packaging, services, and spare parts added another $75.8 million.²

The third tier remains unproven: coater-developer track systems and plasma-enhanced chemical vapor deposition (PECVD) tools, both of which are still undergoing customer evaluation and generate no material revenue.

This structural breakdown explains the company's recent performance. In the June 2026 quarter, the scaling tier drove overall growth: revenue from ECP, furnaces, and other technologies surged 167.7% year over year to $128.5 million, while advanced packaging excluding plating rose 153.3% to $31.4 million.⁷ With non-cleaning lines reaching 54.6% of total revenue, ACM crossed an operational milestone in 2026, transitioning beyond its identity as solely a cleaning equipment specialist.

Why plating is working when polishing did not. ACM's plating platform deposits copper electrochemically—applying the same electrochemical principles as the stress-free polishing tool that failed early in the company's history, but directed at the inverse manufacturing step. The commercial divergence stems less from the underlying physics than from market structure. Plating represented an expanding equipment category with fewer entrenched incumbents, allowing ACM to meet new fab capacity expansions rather than asking customers to replace existing, qualified process standards.

The underlying demand drivers are straightforward. In logic fabrication, larger die dimensions and increasing interconnect layer counts require more copper processing steps per wafer. In memory, high-bandwidth memory (HBM) architectures require vertical DRAM die stacking, where each added layer increases copper plating requirements.⁷ Both trends benefit directly from expanding artificial intelligence infrastructure spending, creating sustained commercial demand for ACM's plating tools.

The panel-level bet. A notable element in ACM's 2026 disclosures is a long-term development initiative in panel-level advanced packaging, an area management began funding over five years ago. As semiconductor dies expand, traditional packaging on round 300-millimeter silicon wafers incurs increasing edge waste. To improve substrate utilization, packaging lines are transitioning toward large rectangular panels, adapting manufacturing concepts from flat-panel display production.

In August 2026, ACM reported securing orders from two advanced packaging customers for its horizontal panel-level plating equipment, supporting both 510×515mm and 310×310mm formats: a commercial production order from an existing mainland Chinese customer and an evaluation unit for a new client elsewhere in Asia.⁷ Management stated that it expects to be among the initial equipment makers delivering horizontal panel-level plating systems across multiple regions, highlighting its horizontal architecture as a key mechanism for maintaining plating uniformity.⁷

While representing a meaningful development step, an initial production order and single evaluation unit mark the start of commercialization rather than a mature revenue driver. ACM's corporate history underscores the need for perspective: early technical innovation does not guarantee market conversion. However, unlike stress-free copper polishing, panel-level packaging does not require displacing an established process standard, as high-volume capacity is still being established without entrenched incumbent equipment.

Vertical furnace progress. Vertical furnace equipment performs high-temperature wafer processing, including low-pressure chemical vapor deposition, oxidation, thermal atomic layer deposition, and annealing. ACM reported an increased furnace revenue contribution in the June 2026 quarter while noting that the product line remains a minor portion of total sales.⁷ Compared to other expansion initiatives, the furnace platform has demonstrated steady commercial adoption, generating recognized revenue rather than remaining confined to customer evaluation.

Track and PECVD assessment. Coater-developer track tools and plasma-enhanced chemical vapor deposition (PECVD) systems represent ACM's most ambitious addressable market expansion, as well as its highest qualification hurdles.

Track equipment applies photoresist and develops wafer patterns, operating physically integrated with lithography scanners—making track qualification among the most sensitive process evaluations in a fab. PECVD deposits thin dielectric and conductive films via plasma energy at reduced thermal budgets. In both categories, ACM faces well-capitalized domestic incumbents, including Kingsemi in track equipment and Piotech in PECVD, with NAURA competing across both market segments.⁴

Management outlined specific progress milestones in August 2026. ACM's proprietary single-chamber, three-track PECVD system completed internal testing at its Lingang R&D facility, and a second unit shipped to a customer in the first quarter of 2026, with full production qualification expected by year-end. Concurrently, a high-throughput KrF track system progressed through fab evaluation, also targeting production qualification by late 2026. Development across track, PECVD, and furnace platforms originated between 2019 and 2021.⁷

These target dates reflect standard industry qualification cycles, where non-cleaning platforms typically require multi-year evaluation periods from initial shipment to formal acceptance. Management's explicit commitment to achieve qualification for both platforms by the end of 2026 provides a clear, near-term test of operational execution.

Until these tools complete customer acceptance, track and PECVD generate no recognized revenue while adding to operating overhead. Research and development expenses reached 13.9% of revenue in the June 2026 quarter, with full-year guidance set between 16% and 18%.⁷ Meanwhile, ACM Shanghai's 2025 gross margin of 48.32% reflected a 1.12 percentage point year-over-year decline, influenced partly by product mix and the cost of funding concurrent development programs prior to commercial revenue.³

Evaluating underlying profitability. A key dimension of ACM Shanghai's financial disclosures involves the composition of its reported earnings.

For the first half of 2026, ACM Shanghai reported revenue of ¥3.718 billion, up 13.87% year over year, and net profit attributable to shareholders of ¥989 million, up 42.14%. However, net profit excluding non-recurring items declined 15.31%. The divergence was driven primarily by fair value gains on equity holdings in listed domestic semiconductor companies—including 华虹 Hua Hong—which generated over ¥470 million in unrealized investment gains during the second quarter alone as share prices appreciated.¹⁹²⁰

The parent entity's consolidated statements reflect a corresponding adjustment: US GAAP non-GAAP net income for the June quarter excluded $69.6 million in unrealized gains on short-term investments.⁷

Consequently, nearly half of ACM Shanghai's first-half headline net profit growth stemmed from investment markups rather than core manufacturing operations, while operating income excluding non-recurring items contracted. Although compliant with standard accounting disclosures, this dynamic indicates that headline net profit expansion outpaced underlying operational earnings growth during the period.

A similar pattern occurred in full-year 2025 results, where reported net profit attributable to shareholders rose 21.05% while net profit excluding non-recurring items increased 10.02%.²¹

Working capital metrics further illustrate the operational capital required for current expansion. Accounts receivable expanded 48.16% in 2025, outstripping revenue growth and reaching a level equivalent to more than double annual net profit. Inventory carrying value surpassed ¥4.8 billion, representing approximately 30% of current assets, while three-year average operating cash flow equaled roughly 17% of current liabilities.²¹ Performance was notably muted in the fourth quarter of 2025, when revenue fell to ¥1.64 billion and net profit dropped 67% year over year to ¥130 million, a decline management attributed to equipment delivery acceptances shifting into early 2026.²¹

While elevated inventory and extended payment cycles are common among semiconductor capital equipment vendors maintaining evaluation tools at customer fabs, these balance sheet metrics highlight the working capital demands and cash conversion dynamics supporting ACM's growth trajectory.

↑ Back to Top

VII. Hamilton Helmer 7 Powers & Porter's 5 Forces Analysis

Frameworks earn their keep when they produce answers that a promotional narrative avoids. Applied to ACM, several provide clear strategic clarity.

Hamilton Helmer's 7 Powers

Process Power — the strongest, and narrower than usually described. ACM's primary advantage is accumulated operational know-how in a domain that resists simple documentation: megasonic acoustic control, fluid bubble dynamics, and handling aggressive hot chemistry without damaging wafer topography or corroding processing chambers. The company's demonstration of its hot SPM technology in May 2026 — achieving fewer than 15 particles at a 15-nanometer threshold with a proprietary nozzle design that prevents acid mist from escaping the chamber and eliminating a recurring maintenance step — illustrates the type of operational capability developed through years of process iteration rather than a single patent filing.⁷

The constraint is scope. Process power in wet cleaning does not transfer automatically to track or thin-film deposition platforms. It represents domain expertise over a specific wafer fabrication step rather than an overarching equipment capability.

Cornered Resource — weak. ACM holds a substantial intellectual property portfolio, and its founder holds more than one hundred patents personally.⁸ However, the company's corporate history demonstrates that patents did not prevent its initial copper polishing technology from failing commercially, while its single-digit global cleaning share indicates that its patent portfolio has not excluded major competitors from the category. Patents function here as defensive intellectual property rather than a cornered resource.

Switching Costs — high where qualified, zero where not. This represents the fundamental structural asymmetry in semiconductor equipment. Once an ACM cleaning process is qualified within a fab's baseline yield model, replacing it requires requalifying a process step that affects downstream layers — a costly, slow, and operationally risky procedure. This inertia is reflected in the company's repeat-order ratio: of more than 1,435 tools delivered since 2009, 1,255 were repeat orders or accepted systems.²

However, high switching costs protect existing positions while creating substantial barriers for new market entry. In coater-developer track and PECVD platforms, ACM faces entrenched incumbents whose process recipes are already integrated into customer fabs. The switching costs that secure ACM's core cleaning franchise act as direct barriers to its platform expansion initiatives.

Scale Economies — moderate and improving. ACM's Lingang facility brought its first manufacturing building into volume production by August 2026, with a second facility planned to open later in the year; management estimates the combined site can support up to $3 billion in annual output.⁷ That infrastructure provides operational leverage for a business operating near $1.1 billion in annual revenue. Offsetting this manufacturing capacity, ACM employed 2,513 people at the end of 2025, including 1,228 research and development personnel — representing nearly half of its total workforce.² An engineering-intensive headcount structure creates a substantial fixed operating expense that fixed-asset scale does not quickly dilute.

Counter-Positioning — weak, with one exception. Established equipment makers like SCREEN Holdings and Lam Research face no structural disincentives to building hybrid batch-plus-single-wafer tools or phase-shifted megasonic systems, as adopting those architectures would not cannibalize an existing business model. The partial exception lies in geopolitical positioning: foreign incumbents cannot sell freely into Entity-Listed Chinese fabricators or restructure their operations to bypass export controls. That competitive isolation stems from trade policy rather than corporate strategy, leaving it subject to regulatory change.

Branding and Network Economies — not material. Semiconductor fabricators evaluate equipment based on process qualification data and yield economics rather than brand perception, and individual tool installations generate no network effects across customer fabs.

Porter's 5 Forces

Bargaining power of buyers — high, though measurably less concentrated than a year ago. State-backed semiconductor fabs operate formal multi-sourcing procurement programs, require multi-month on-site evaluations prior to commercial acceptance, and maintain active programs to qualify domestic alternatives. While lower customer concentration in the first half of 2026 reduced single-customer revenue risk, it has not altered the structural bargaining leverage held by major fab operators.⁷

Bargaining power of suppliers — moderate to high, and currently tightening. ACM relies on specialized mechanical assemblies, precision fluid delivery components, robotics, and motion subsystems, with key suppliers concentrated in Japan and South Korea. Management reported component lead times extending from four months to six, prompting the company to pre-purchase raw materials to secure production schedules.⁷ Asked whether rising component costs could be passed through to buyers, CEO David Wang stated that ACM was not raising tool prices and that suppliers had generally maintained prices while extending delivery schedules.⁷ Suppliers that delay component deliveries retain operational leverage over equipment build schedules and revenue timing, even if gross margins remain temporarily insulated. An inventory balance of $783.1 million, including $406.1 million in raw materials, reflects the working capital required to buffer these supply chain delays.⁷

Threat of rivals — high and rising in the categories ACM is entering. The domestic competitive landscape features well-capitalized, publicly traded equipment vendors, including market leader NAURA Technology Group, which maintains a market capitalization substantially larger than ACM Shanghai.⁴ In international markets, ACM operates as a challenger against four established incumbents that collectively control over 90% of the global cleaning equipment market.⁶

Threat of substitutes — low. Wet chemical processing remains indispensable across dozens of semiconductor manufacturing steps, with no alternative technology positioned to replace liquid chemistry. Supercritical carbon dioxide and advanced dry processing techniques target specific residue removal and drying steps, serving as complementary processes rather than direct replacements. Wet cleaning represents the single equipment category where substitution risk remains minimal. Furthermore, management expects the category's market size to expand: Wang has argued that wafer cleaning could increase from its historical 5% to 7% share of total fab equipment spending toward 10% as device feature sizes shrink and particle specifications tighten.⁷ That projection reflects management expectations rather than an established industry consensus.

Threat of new entrants — low in cleaning, already realized elsewhere. Entry barriers in wafer cleaning are defined by process qualification cycles rather than capital requirements. However, potential competitive threats have already materialized in adjacent equipment categories: the domestic market leaders ACM now confronts in track and PECVD did not exist as capable competitors a decade ago.

↑ Back to Top

VIII. Investor Playbook & Earnings Call Analysis

A brief exchange on ACM's August 2026 earnings call highlighted an important pending capital structure transition. When an analyst inquired about an update on "the Shanghai listing," CFO Mark McKechnie clarified, "Hong Kong, right?" CEO David Wang then declined to elaborate beyond confirming that the company had announced its intention in April and that further details would be disclosed at a later date.⁷

The exchange points to a significant pending development in ACM's corporate structure. On April 17, 2026, ACM Shanghai notified the Shanghai Stock Exchange of its proposal to issue overseas-listed shares and seek a listing on the Stock Exchange of Hong Kong.²² The board approved the plan on May 26, 2026, authorizing an H-share issuance representing up to 7% of total enlarged share capital. Management cited international expansion, capital base strengthening, and global talent retention as primary objectives, while explicitly noting substantial regulatory and approval uncertainties.²²

For an enterprise expanding internationally amid heightened US trade restrictions, a Hong Kong listing offers clear strategic logic: it broadens the offshore investor base and creates an international equity currency without requiring the US parent company to sell additional shares in its Shanghai subsidiary. Mechanically, however, the proposed issuance would dilute existing shareholders by up to 7% while adding a third listed entity to an already complex corporate footprint. Whether the transaction proceeds, and under what final terms, remains unresolved.

Reading management's record on promises. Evaluating executive execution reveals distinct strengths alongside ongoing operational challenges. On process technology, leadership has demonstrated foresight: management committed to panel-level packaging research and development more than five years before market adoption accelerated, a strategic bet validated by initial commercial and evaluation orders secured in 2026.⁷ Similarly, pivoting early from stress-free copper polishing to wet cleaning successfully redirected the company toward a viable commercial market.

The company's financial guidance record, however, shows greater variability. For 2025, management initially projected ACM Shanghai revenue of up to ¥7.1 billion, whereas the subsidiary ultimately reported ¥6.786 billion.²³³ While this outcome represented a minor shortfall relative to the upper bound rather than a midpoint miss, it coincided with a steep fourth-quarter profit drop as equipment acceptances shifted into the following year.²¹

For full-year 2026, management adjusted guidance upward, raising the lower bound of its revenue range in August following strong first-half performance.⁷ This pattern reflects a practice of issuing broad initial target ranges and refining them as operational delivery unfolds, providing wider variance bands until late in the fiscal year.

Execution timelines for adjacent product platforms represent a more evident area of delay. Development of coater-developer track and PECVD tools began between 2019 and 2021, yet both product lines remain in customer evaluation.⁷ Despite repeated optimistic commentary regarding near-term commercialization, neither platform has reached recognized volume revenue. Management's explicit commitment to achieve production qualification for both platforms by late 2026 establishes a concrete benchmark for assessing platform execution.

What analysts push on, and how the answers land. Three recurring topics dominate investor inquiries.

The first topic is gross margin trajectory. ACM maintains a long-term gross margin target of 42% to 48% and reported 46.0% for the June 2026 quarter, compared to 48.7% in the prior-year period.⁷ Addressing questions regarding component cost inflation, McKechnie stated that pre-purchased raw material inventories insulated near-term profitability, mitigating cost pressures.⁷ Notably, management's target range implies that as product mix shifts toward lower-margin adjacent equipment, gross margins may trend toward the lower end of the 42% to 48% band rather than sustaining high-forties levels over the long term.

The second topic concerns the gap between order growth and recognized revenue. When asked why second-half financial estimates might require moderation despite first-half outperformance, Wang highlighted supply chain constraints delaying equipment shipments, while McKechnie noted that a portion of 2026 orders would shift into 2027 revenue.⁷ This acknowledgement confirms that despite a 105% year-over-year surge in first-half orders, supply chain lead times will prevent immediate conversion into near-term recognized sales.

The third topic focuses on international market expansion. Management projects more than 20 tools installed across approximately 10 customer fabs in five countries outside mainland China by the end of 2026, supported by active deployments in Singapore and North America and the opening of an Oregon demonstration center.⁷ Relative to the company's long-term international revenue target of $1.5 billion, an initial base of 20 installed tools highlights the early phase of overseas expansion, reflecting Wang's observation that global deployment remains in its initial stages.⁷

The credibility summary. ACM's corporate record reflects strong core process engineering capabilities alongside mixed execution in commercializing adjacent equipment lines. Financial communication remains relatively transparent, with management disclosing unfavorable segment results directly—such as highlighting the 14.2% year-over-year decline in June-quarter cleaning revenue in prepared comments rather than omitting it.⁷ However, achieving long-term growth targets depends on demonstrating sustained commercial acceptance across newly introduced equipment platforms.

↑ Back to Top

IX. Bull vs. Bear Stress Test & Key KPIs

Strip away the narrative and the argument over ACM Research (Shanghai) reduces to a single disagreement: whether the company is a differentiated process-technology firm that happens to sell mostly in China, or a Chinese localization beneficiary that happens to own some good technology. The two readings produce very different conclusions about what happens when the indigenization wave matures.

The bull case, at its strongest.

The first pillar is that the core franchise is real and was established through technical execution. ACM won its position in a category dominated by four foreign incumbents by inventing acoustic methods that solved a genuine process dilemma, and it proved them at a tier-one Korean memory customer before Chinese fabs had the volume to serve as global reference accounts. The 48.32% gross margin ACM Shanghai earned in 2025 is not the margin profile of a commodity local substitute.³

The second pillar is that the platform transition is already under way beyond the high-profile experimental lines. Plating and advanced packaging grew by triple digits in the June quarter and, together with vertical furnaces, pushed the non-cleaning business past half of total revenue for the first time.⁷ The plating position benefits from two of the most durable structural trends in semiconductors — rising interconnect layer counts in logic and vertical DRAM stacking for high-bandwidth memory — while initial panel-level plating orders give ACM an early position in a format the industry is only beginning to adopt.⁷

The third pillar is capacity and balance sheet strength. The Lingang facilities can support up to $3 billion in annual output against a business running near $1.1 billion, meaning near-term expansion does not require a new capital expenditure cycle.⁷ The group held $1.36 billion in gross cash and $1.0 billion in net cash at the end of June 2026, and ACM Shanghai has raised onshore capital — through its ¥3.685 billion IPO and a subsequent placement priced at ¥116.11 per share — without requiring capital from its parent company.⁷[^17]¹⁷

The fourth pillar is the expansion of the addressable market. Industry estimates from Frost & Sullivan in June 2026 projected that the global semiconductor equipment market exceeded $140 billion in 2025 and would surpass $200 billion by 2029, with mainland China representing over $50 billion in 2025 and heading beyond $80 billion.⁷ Even modest share gains against that expanding baseline support sustained revenue growth.

The bear case, at its strongest.

The most immediate challenge is that the core cleaning line is contracting. Cleaning revenue fell 14.2% in the June quarter.⁷ Management's explanation of a product-line transition is plausible and supported by finished-goods inventory balances, but qualified domestic competitors — including NAURA, Kingsemi, and PNC Process Systems — are well-funded and targeting the same customer fabs, leaving ACM's 23% domestic cleaning market share vulnerable.⁶⁴ If the decline persists into 2027, the transition narrative loses credibility, indicating market share erosion in ACM's highest-margin category.

The second concern is earnings quality. Net profit excluding non-recurring items fell 15.31% in the first half of 2026 while headline attributable net profit rose 42.14%, a divergence driven primarily by mark-to-market gains on equity holdings in listed domestic semiconductor companies.¹⁹²⁰ Investors buying 688082 for exposure to semiconductor equipment are simultaneously taking on a leveraged position in the Chinese semiconductor equity market — an exposure that fluctuates with broader equity markets rather than equipment competitiveness.

The third issue involves working capital efficiency. Accounts receivable expanded significantly faster than revenue in 2025, inventory carrying value exceeded ¥4.8 billion, and three-year average operating cash flow relative to current liabilities remained modest.²¹ Growth driven by deploying evaluation tools that customers have not yet accepted or paid for carries substantial working capital requirements.

The fourth constraint stems from export control policy. ACM Shanghai remains designated on the US Entity List.¹ The company's long-term revenue target of $4 billion assumes $1.5 billion generated outside mainland China.⁷ These goals exist in direct tension with export restrictions set by foreign regulators. Meanwhile, filings from the US parent entity highlight ongoing regulatory risks, including PCAOB audit-access compliance under the Holding Foreign Companies Accountable Act.²

The fifth challenge centers on track and PECVD equipment. Bullish arguments treat these adjacent lines as low-cost optionality. In practice, they require substantial research and development funding — guided at 16% to 18% of sales — while ACM's corporate history includes an initial decade-long effort that failed to displace established process standards.⁷⁹ Historical conversion rates for challengers entering established tool categories remain low, a dynamic ACM encountered during its early corporate history.

Weighing the two. The empirical evidence supports a calibrated version of the bull case rather than either extreme narrative.

ACM's process differentiation in cleaning is supported by its commercial record — a tier-one Korean qualification predating domestic localization mandates, a repeat-order ratio exceeding 85% of total tool deliveries since 2009, and gross margins well above commodity equipment levels.²³ These factors enabled ACM to capture domestic market share ahead of lower-tier competitors.

However, the rapid acceleration of revenue growth was heavily influenced by trade policy. Fifteen years of organic development yielded a 6.6% global cleaning market share; four years of trade restrictions coincided with a nearly sixfold revenue expansion.⁶¹⁵² Commercial execution in international markets without trade-driven tailwinds remains unproven, particularly under Entity List restrictions.

Platform expansion efforts require separate evaluation. The electrochemical plating business represents a scaled operating segment with an installed base of 2,000 chambers and early positions in panel-level packaging.⁷ Furnace equipment shows steady commercial progress. By contrast, track and PECVD platforms remain speculative, with management's target of securing production qualification by late 2026 serving as the primary test of execution.

An activist's brief. A skeptical investor would focus on four core questions. First, why does a company generating substantial reported profits hold large equity positions in listed peers rather than returning capital to shareholders or reinvesting in core operations? Second, why have accounts receivable consistently outpaced revenue growth, and what does this pattern indicate about customer payment terms during China's fab expansion? Third, what is the contingency plan if track and PECVD platforms fail to secure production qualification by year-end, and at what point will research spending on these initiatives be reassessed? Fourth, what is the parent company's long-term strategy for its equity stake, given that it has twice reduced its ownership position while stating it has no immediate plans for further sales?¹⁸⁷ The proposed Hong Kong listing adds potential equity dilution to these governance considerations.²²

The three metrics that will settle the argument.

1. Single-wafer cleaning revenue, year over year. This metric provides a direct test of whether the June quarter revenue decline represented a temporary product transition or structural market erosion. Key data to monitor include the conversion of high-temperature SPM evaluation shipments into recognized revenue: management estimates SPM accounts for roughly one-third of the total cleaning equipment market, has generated minimal recognized revenue to date, and expects to ship over twenty tools during 2026, with revenue recognized upon customer acceptance in subsequent quarters.⁷ A rebound in cleaning revenue during 2027 would validate the transition thesis, whereas continued weakness would indicate market share loss to domestic competitors and pressure on overall gross margins.

2. Non-cleaning revenue contribution, and whether track and PECVD actually qualify. Non-cleaning equipment has reached over half of total product mix, driven by plating and packaging systems.⁷ The immediate operational test is binary: management committed to achieving full production qualification for both the KrF track tool and the PECVD platform by the end of 2026.⁷ Successful qualification would convert these evaluation programs into revenue-generating lines and expand the company's addressable market. Continued delays would leave these platforms in the same status as early stress-free polishing — technically viable, commercially unproven, and consuming R&D resources.

3. Recurring profit excluding non-recurring items, read against the gross margin band. This metric evaluates underlying operational profitability. Headline attributable net profit has been elevated by mark-to-market investment gains, while recurring operating profit grew at a slower rate and contracted in the most recent half-year period.¹⁹²⁰²¹ Evaluating recurring profit alongside gross margins relative to management's long-term target band of 42% to 48% will indicate whether product mix diversification maintains operational profitability.⁷ Recurring profit growth aligned with revenue expansion and gross margins holding in the upper half of the target range would confirm a successful platform transition, whereas flat recurring profit alongside margins trending toward the lower bound would indicate that revenue diversification is diluting profitability.

↑ Back to Top

X. Epilogue & Strategic Lessons

A single image captures the company's founding tension: David Wang in a Fremont office in 1999, holding a wafer polished without mechanical stress, explaining to a fab engineer why the process was technically superior—and the engineer explaining why the fab would stick with conventional CMP anyway.

Everything ACM Research built afterwards grew from how Wang responded to that rejection. He did not abandon the technology; the stress-free polishing line remains folded inside a minor disclosure category at the Shanghai subsidiary nearly three decades later.¹⁰ Instead, he recognized that superior process physics does not guarantee a commercial market. He set out to find a manufacturing bottleneck where incumbent solutions were not entrenched, in a market where local supply chain presence would eventually carry strategic weight.

The strategic shift—relocating engineering and manufacturing operations to Zhangjiang in 2005 and committing to a 100,000-square-meter facility in Lingang beginning in 2020—proved decisive.¹¹²⁴ Wang did not foresee the 2022 trade restrictions. Instead, he built an operating base where cost-effective engineering talent and domestic supply chain requirements converged. When trade policy accelerated local procurement, ACM had already established qualified equipment in mainland fabs.

The lessons the record actually supports.

The first lesson is that technical performance in semiconductor capital equipment is merely an entry requirement. Commercial adoption depends on process qualification—a risk calculation made by fab managers accountable for yield stability. ACM's history demonstrates both outcomes: an initial copper polishing technology that failed to gain adoption, followed by a megasonic cleaning process that secured tier-one qualification. That contrasting record explains why its expansion into coater-developer tracks and thin-film deposition requires rigorous commercial validation rather than optimistic projection.

The second lesson is that product adjacency is not universally transferable. ACM's core process capabilities in acoustic cleaning transferred effectively into wet etching and electrochemical plating, sharing chemical systems, fluid dynamics, and chamber architectures. By contrast, those capabilities do not extend to plasma thin-film deposition or track photoresist processing. While plating scaled rapidly into a major revenue contributor, track and PECVD platforms have required five to seven years of development and remain in customer evaluation. Process overlap that appears logical on a product roadmap faces separate engineering realities on the fab floor.

The third lesson involves distinguishing between policy tailwinds and structural moats. Trade restrictions provided ACM with access to domestic fab customers that would have taken years of conventional qualification to secure. However, regulatory tailwinds can change, and US authorities have designated the operating subsidiary on the Entity List. The defensible portion of ACM's business rests on its qualified installed base and repeat-order rates, whereas growth dependent on domestic procurement mandates remains subject to fab expansion cycles and policy shifts.

The fourth lesson highlights the importance of evaluating core operating earnings separate from investment markups. The first half of 2026 showed core operating income declining while headline net profit expanded, a divergence driven by unrealized gains on equity holdings in listed domestic semiconductor peers.¹⁹²⁰ Investors assessing the equipment franchise must differentiate underlying manufacturing profitability from mark-to-market portfolio fluctuations.

ACM has built a rare enterprise: commercializing proprietary process technology, surviving an extended development phase without volume sales, finding market adoption in specialized cleaning, and expanding during a major domestic supply-chain transition. However, proving that its platform expansion can succeed outside mainland China—or that adjacent tool lines will mirror plating rather than early copper polishing—remains an open question. Operational benchmarks are now established: production qualification targets for track and PECVD platforms by late 2026, alongside an international deployment target of more than twenty tools. Those milestones will determine whether ACM can translate domestic market presence into a broader, globally competitive equipment platform.