Nearfield Instruments

1.1 Identity, spin-out roots, and product definition

Nearfield Instruments is now best understood as a Rotterdam-based semiconductor metrology company with unusually strong research lineage and unusually thin public financial disclosure. TNO and TNO Ventures establish the company's 2016 spin-out origin, while the company's own pages anchor the current product identity around QUADRA and AUDIRA. Those products are not generic inspection tools: QUADRA is presented as a high-throughput atomic-force-microscopy platform for on-device 3D process control, while AUDIRA extends the platform into buried-feature and void inspection through subsurface acoustic readout. The official Series D release and the TNO coverage both position Nearfield as part of the critical infrastructure required for AI-era chip manufacturing rather than a narrow niche supplier. That framing matters because it explains why investors funded the company at a unicorn valuation before the company disclosed conventional software-style KPIs such as ARR, revenue run rate, or customer count. The identity picture is therefore strong on technical mission and weak on monetization detail. The public profile is also unusually coherent for a private European hardware business because the same origin story, product framing, and advanced-node use cases recur across company, investor, and institutional partner sources. That consistency materially reduces ambiguity about the company's strategic lane even as operating metrics remain sparse in public materials still.[CO001, CO002, CO003, CO004, CO005, CO006]

1.2 Leadership, capital base, and governance dependence

The public record points to a founder-led company whose credibility is driven by technical continuity and investor validation rather than transparent governance disclosure. Hamed Sadeghian remains the central public face of the business across Nearfield, TNO, and partner releases, and the careers surface shows a company still hiring deeply into engineering and operations instead of shifting into a low-intensity commercial model. The June 2026 Series D round materially changed the capital base: Fidelity led, QIA joined, and an existing coalition spanning Temasek, Walden Catalyst, Innovation Industries, M&G, Invest-NL, TNO Ventures, and ING doubled down. The use-of-proceeds language is specific about production, customer support, and applications centers, which fits a capital-equipment company that must scale hardware, service, and collaborative development simultaneously. What remains weak is formal governance visibility: public sources do not provide current board composition, ownership percentages, or voting control, so the chapter treats governance dependence as an unresolved diligence area rather than a verified strength. The hiring mix also suggests that module industrialization, supplier coordination, and export-control compliance are core scaling tasks rather than peripheral support functions. That combination is typical of a company moving from breakthrough platform validation into repeatable factory execution.[CO006, CO009, CO010, CO011, CO012, CO013]

1.3 Milestones, external proof, and remaining disclosure gaps

Nearfield's milestone chronology is unusually compelling for a private European hardware scale-up. TNO documents the 2020 shipment of the first QUADRA system to Samsung Semiconductor, the High Tech Campus article states that Samsung was a customer by mid-2024, and the November 2025 imec project put QUADRA into one of the world's most credible advanced-node R&D environments. The 2026 Series D then added the strongest possible financial signal short of an IPO. At the same time, the negative space in the record is real. Neither the company nor third parties disclose current revenue, installed-base count, customer concentration, margin structure, or clear board governance. Verdict's IPO story and Business News Today's KLA framing make that tension explicit: Nearfield may have crossed the credibility threshold on technology and fundraising, but public evidence is still insufficient to underwrite operating maturity at the same level of confidence. The result is a chapter-level view that is bullish on strategic relevance and cautious on commercial completeness. In practical diligence terms, the next step after reading public materials is not to challenge whether the company matters, but to narrow how much of that importance has already translated into durable customer economics, repeat orders, and governable scaling discipline.[CO018, CO019, CO020, CO021, CO022, CO023]

1.4 Exhibits

2.1 Market boundary and sizing discipline

The right market boundary for Nearfield is narrower than generic wafer-fab equipment and broader than stand-alone AFM. Official product pages from KLA, Applied Materials, ASML, Nova, and Onto all describe a process-control layer that spans metrology, wafer inspection, defect review, overlay, materials characterization, and software used from R&D through high-volume manufacturing. KLA's filing and Applied Materials' process-control framing make the operational need vivid: manufacturers use metrology and inspection to catch drift before it damages yield, profitability, or production-ramp speed. That is why process control deserves its own market lens rather than being treated as a footnote inside lithography or deposition capex. Public market estimates confirm that this layer is already substantial, but they do not line up cleanly enough to justify a single precision TAM. Mordor Intelligence places the semiconductor metrology-and-inspection system market at $13.03 billion in 2025 and $16.95 billion in 2030. Yole's 2026 report summary, cited by DQIndia, says the market exceeded $18 billion in 2025 and represented roughly 14% of wafer-fab-equipment spending. Those are directionally consistent on growth, but they clearly embed different boundaries, likely including different mixes of software, packaging tools, or adjacent inspection modalities. The correct diligence move is therefore to preserve multiple lenses and explicitly state that no public source cleanly isolates the AFM-plus-buried-feature subset that matters most for Nearfield.[CM001, CM002, CM003, CM004, CM005, CM006]

2.2 Buyers, users, and the adoption path

The buyer map is also more specialized than a broad semiconductor-equipment screen suggests. The most relevant users are process-integration, yield-engineering, and metrology teams inside leading-edge logic foundries, memory producers, advanced-packaging or OSAT operations, and elite R&D lines such as imec-like development environments. Budget ownership usually sits with process-control or fab-capital committees rather than product-design groups, because the tool's value is measured through yield protection, line qualification speed, and defect escape reduction. Onto's public materials and quarterly results reinforce that reality by tying current customer wins to 2.5D logic, HBM, and gate-all-around metrology rather than to generic fab expansion. The adoption path usually runs from technical problem definition to pilot qualification and only then into fleet standardization. That path matters for Nearfield because the category is full of status-quo substitutes, not an empty greenfield. Buyers can often respond to a new measurement problem with optical CD, ellipsometry, e-beam inspection, off-line AFM, or more intense sampling using incumbent process-control stacks before they grant a new vendor tool-of-record status. Public pages from Bruker, SENTECH, Onto, and ASML/HMI make this substitute set visible. Nearfield's opportunity is therefore not simply that metrology demand is rising; it is that some next-generation use cases may require 3D or buried-feature measurement where legacy optical or inspection approaches lose efficiency or accuracy.[CM009, CM010, CM011, CM012, CM013, CM014]

2.3 Growth drivers, constraints, and what remains unknowable publicly

The strongest demand drivers are complexity-driven rather than unit-driven. IRDS highlights AI and HPC workloads, chiplets, advanced packaging, hybrid bonding, backside power delivery, and HBM as technologies that require new metrology approaches. DQIndia's summary of Yole pushes in the same direction, arguing that metrology and inspection grew faster because AI, HBM, advanced packaging, hybrid bonding, and High-NA EUV make process control a strategic bottleneck rather than a support function. KLA's recent results add a live commercial check on that thesis: management says the AI-infrastructure buildout is benefiting foundry/logic, memory, advanced packaging, and services simultaneously. In other words, the category tailwind is not hypothetical. The main constraints are equally real. Dutch export-control changes in 2024 and 2025 explicitly broadened authorisation requirements for advanced semiconductor manufacturing equipment and then for specific measuring and inspection equipment, which matters disproportionately to a Dutch supplier such as Nearfield. The market is also highly incumbent-shaped: KLA, ASML/HMI, Onto, Nova, Applied already occupy adjacent workflow positions with software, service, and installed-base credibility. Public sources therefore support a strong category-growth thesis but an incomplete company-capture thesis. The missing diligence asks are the ones that matter most to underwriting: exact AFM/subsurface SAM, buyer-by-buyer budget share, deployment-cycle length, and the share of demand that truly escapes incumbent substitute tools.[CM025, CM026, CM027, CM028, CM029, CM030]

2.4 Exhibits

3.1 Landscape structure and positioning logic

The competitive set around Nearfield is layered rather than singular. At the top sit process-control incumbents such as KLA, ASML/HMI, Applied Materials, Nova, Onto, and Hitachi High-Tech, all of which present broad portfolios spanning inspection, metrology, software, or adjacent yield-management surfaces. These are the companies a leading-edge fab already knows, budgets for, and often integrates across multiple steps. External mapping sources reinforce that framing: CB Insights places KLA, Oxford Instruments, SENTECH, and Hitachi High-Tech among Nearfield's alternatives, which means third parties also view the company through a cross-section of direct, incumbent, and adjacent vendor classes rather than as a category of one. That matters because Nearfield's public differentiation is not portfolio breadth. The clearest differentiation in public materials is specialized 3D or subsurface process control rooted in scanning-probe-style measurement, while most incumbents sell much broader optical, e-beam, defect-inspection, and software-led process-control stacks. In other words, Nearfield is likely to win by being uniquely useful in a narrow set of hard measurement problems, not by matching the incumbents feature for feature. A realistic competitor map therefore has to include both the giants that can bundle overlapping capabilities and the narrower research or AFM vendors that give buyers a cheaper or less risky substitute for some use cases.[CP001, CP002, CP003, CP004, CP005, CP006]

3.2 Capability breadth, packaging, and commercial contrast

On capability breadth, the incumbents look nothing like a young specialist. KLA's 10-K describes a process-control portfolio that spans inspection, metrology, software, and services from R&D to high-volume manufacturing. ASML's HMI business is pushing multi-beam e-beam inspection into 3 nm volume manufacturing. Onto markets advanced metrology across logic, memory, and advanced packaging and has already announced fresh customer wins in HBM and gate-all-around workflows. Nova and Applied likewise frame their offerings around integrated dimensional, materials, and process-control measurement rather than one narrow modality. Nearfield may have sharper differentiation in a smaller set of measurements, but the public competitive burden is obvious: every account pitch happens against vendors that can surround the same buyer with broader tool, service, and analytics coverage. The commercial comparison is asymmetric as well. Public competitors disclose quarterly results, long-run revenue histories, and public-market valuation signals that make scale legible. Nearfield does not. Even pricing is opaque across much of the category: product pages generally present quote-led enterprise sales motions, contact paths, or solution marketing rather than list pricing. That means the practical pricing comparison is less about posted dollar tags and more about packaging model, integration burden, service dependency, and whether the tool sits as a single-purpose add-on or as part of a broader process-control standard. Nearfield's opportunity is that buyers may still pay for a narrow tool when the measurement pain is acute; the challenge is that most public rivals can tie their tool pitch to a wider platform story.[CP004, CP005, CP006, CP007, CP008, CP009]

3.3 Switching costs, incumbency power, and moat durability

The deepest moat question is not whether Nearfield has interesting technology; it is whether that technology can become embedded deeply enough in the fab workflow to overcome incumbent account control. KLA's filings, ASML/HMI's installed-lineage story, and Hitachi's process-control framing all point to the same structural fact: once a vendor becomes part of the yield-management loop from pilot through HVM, switching costs rise because uptime, data continuity, service support, and adjacent-tool compatibility all matter. That creates a competitive environment where specialists must prove not just a better measurement, but a better operational outcome than the incumbent workaround. The public evidence suggests Nearfield's moat is strongest in use cases where buried-feature, 3D, or AFM-derived data reveals something the broad platforms cannot deliver efficiently. Its moat is weakest wherever buyers can solve the problem through optical CD, e-beam inspection, thin-film metrology, or simply more intensive use of existing process-control stacks. Advanced packaging, HBM, and gate-all-around should create more opportunities for specialists, but those same drivers mobilize large incumbents with more distribution power, larger service organizations, and more public proof. The net view is therefore balanced: Nearfield appears differentiated, but the durability of that differentiation depends on precise workflow wedges that the public record still does not quantify well enough.[CP027, CP029, CP030, CP031, CP032, CP033]

3.4 Exhibits

4.1 Revenue model, monetization path, and what is still hidden

Public evidence supports a hardware-led revenue story, but only in outline. Nearfield’s official materials describe QUADRA and AUDIRA as capital-equipment platforms sold into advanced semiconductor process-control workflows, not as software seats or transaction rails. That matters because it implies that recognized revenue should be tied primarily to system sales, qualification, and deployment work, with service, applications, and support layered around the installed base. The 2026 Series D release strengthens that interpretation: management did not say the capital would fund customer-acquisition spend in a low-touch model; it said the money would expand production capacity, shorten delivery times, broaden customer support, and build applications centers. Those are the fingerprints of a scale-up moving deeper into industrial delivery. At the same time, the public record stops short exactly where underwriting would begin. There is no disclosed revenue run rate, no public service-share split, no realized pricing, no disclosed gross margin, and no contract-term data. The result is a chapter in which the monetization mechanism is visible, but the monetization quality is not. Investors can see that Nearfield likely has multiple revenue-bearing activities around tool deployment; they still cannot tell how much of reported value is recurring, how much is one-time system recognition, or how quickly service economics improve after installation.[CI001, CI002, CI004, CI005, CI006, CI007]

4.2 Capital intensity, unit-economics proxies, and comp brackets

The public hiring trail makes Nearfield’s economic profile look like a deep-tech equipment business with real industrial overhead. The technical project lead role spans engineering, manufacturing, procurement, and quality, while the optics engineer role owns critical modules from concept through volume production. That means the cash model almost certainly includes supplier management, precision components, mechatronics, applications engineering, and field support, not just front-end research. Because Nearfield does not publish direct unit-economics data, public-company peers are the best available bracket. KLA shows what scaled process control can look like at the high end: more than $12 billion of annual revenue, gross margins a little above 60%, and multibillion-dollar free cash generation. Onto provides a smaller but still meaningful reference with roughly $292 million of Q1 2026 revenue and gross margins around 50% GAAP / 55.7% non-GAAP. Nova’s sub-$1 billion revenue base provides another scale marker for a more specialized metrology vendor. None of those figures is a proxy for Nearfield itself, but together they tell the right story: if a process-control vendor wins technically and operationally, margins can become attractive at scale. What remains unknowable publicly is where Nearfield sits on that curve today — before or after the point where installations, service burden, and manufacturing complexity start to turn into durable contribution margin.[CI010, CI011, CI012, CI018, CI019, CI020]

4.3 Capital adequacy, disclosure limits, and the investability verdict

The good news is that capital adequacy improved in a concrete way in June 2026. Nearfield raised one of the largest deep-tech rounds in the Netherlands, and both the company and TNO describe the proceeds in operational terms that fit the next phase of scaling: production capacity, delivery, support, and applications infrastructure. The bad news is that those same statements highlight why the company is still difficult to underwrite from public materials alone. Management is clearly preparing to spend, yet investors cannot calculate how fast the money burns because there is no public cash balance, no runway disclosure, no gross-margin history, no backlog number, and no visibility into service share or customer concentration. The adverse reading is therefore not that Nearfield lacks traction; it is that the company’s public profile still outruns its financial disclosure. Public-market comparables also cut both ways. Category leaders such as KLA and ASML can command double-digit sales multiples, but instrumentation-like businesses can trade dramatically lower, which means Nearfield’s own $1.6 billion valuation could prove cheap or expensive depending on revenue quality that outsiders still cannot see. The practical diligence conclusion is straightforward: funding risk looks far lower after Series D, but underwriting risk remains high until management discloses actual revenue, gross margin, burn, and repeat-order behavior.[CI001, CI002, CI003, CI027, CI028, CI029]

4.4 Exhibits

5.1 Product definition and customer workflow

Nearfield’s product story is much clearer than its public financial story. The company is selling metrology systems that sit directly in advanced semiconductor process-control workflows, not generic laboratory tools and not software abstractions. QUADRA is the core inline 3D scanning-probe platform, while AUDIRA extends the stack into subsurface and buried-feature inspection. Around that base platform, Nearfield is publishing a visible capability-upgrade ladder: Lightning Mode for faster imaging throughput, Sidewall Imaging Mode for more complete 3D access in extreme high-aspect-ratio structures, and an imec-backed roadmap for high-NA EUV, CFET, hybrid bonding, and buried-power-rail metrology. That means customers are not buying a frozen box; they are buying into a measurement architecture that can add modes and applications over time. The strongest public proof is that these are customer-workflow tools, not merely technical demos. Nearfield repeatedly frames them around process monitoring, yield control, full-wafer coverage, and inline deployment. The weakest public area is still commercial-operational detail: there is no public installed-base count, no uptime disclosure, and no public map of which capabilities are sold as standard features versus separately monetized options.[CE001, CE002, CE003, CE004, CE005, CE006]

5.2 Architecture, dependencies, and maturity evidence

The technical architecture that emerges from public sources is unusually specific for a private hardware company. The academy technology page ties Nearfield’s inline measurement capability to FFTP trajectory planning, advanced mechatronics, long- and short-range positioning stages, and six-degree-of-freedom vibration isolation. The new Sidewall Imaging Mode adds sidewall force sensing and sensor fusion, while AUDIRA mixes acoustic microscopy with AFM to inspect buried structures. Independent technical literature sharpens the use cases: the DSPE paper maps QUADRA into hybrid bonding, buried-power-rail, and gate-all-around manufacturing problems, and the imec project pushes the platform into high-NA EUV resist work, CFET sidewalls, and heterogeneous integration. Maturity evidence also exists, though unevenly. The about page says QUADRA was deployed in HVM by late 2023 and TNO records a 2020 shipment to Samsung. At the same time, the dependency picture is still incomplete. Nearfield clearly depends on high-precision mechanics, control software, applications engineering, and regional field support, but it does not publicly name critical suppliers, contract manufacturers, or uptime benchmarks.[CE012, CE013, CE014, CE015, CE016, CE017]

5.3 Support footprint, trust gaps, and the technical verdict

Operationally, Nearfield looks more mature than a simple R&D spin-out. The management roster now includes explicit finance, sales, operations, system-engineering, and program-management leaders, while contact surfaces list offices or representation across the Netherlands, Japan, South Korea, the United States, Singapore, and Taiwan. The hiring surfaces reinforce that the company is still in build mode: it is recruiting for volume-production optics ownership and cross-functional industrialisation leadership rather than only pure research. That is a strong signal for deployment intent and regional support readiness. The trust side is where the public record thins out. There are no cited uptime or SLA numbers, no public quality-certificate list, no explicit software-security control stack for customers, and no public supplier map. Just as important, there is no public field-failure history, no quantitative support-response disclosure, and no third-party audit packet that would let an outside diligence team stress-test operational readiness from documents alone. None of those gaps disproves the product; they simply mean the technical diligence is not finished. The correct product conclusion is therefore balanced. Nearfield’s public evidence is strong on modality, roadmap specificity, and deployment relevance in difficult next-generation nodes. It is still incomplete on operational reliability, certification depth, and supply-side resilience, which are exactly the areas a buyer would probe before making the company a durable tool-of-record standard. Those missing layers matter because metrology vendors are judged not only on image quality, but on whether they can stay inside line control loops without creating operational drag for the fab, especially during qualification windows, line excursions, and fab changeovers. That operational threshold is decisive.[CE019, CE020, CE021, CE022, CE023, CE024]

5.4 Exhibits

6.1 Customer segments and named proof

Nearfield’s customer evidence is narrow but real. Publicly named proof clusters in exactly the kinds of accounts one would expect for a frontier process-control company: a leading-edge fab customer in Samsung, a development and validation partner in imec, and an unnamed U.S. customer that had already placed multiple orders by early 2025. That is not a broad-base customer story, but it is a serious one. The strongest signal is Samsung. Independent 2024 reporting and TNO’s retrospective both say the company shipped QUADRA into Samsung’s environment, and multiple sources still describe Samsung as a customer rather than a one-time evaluator. Imec is different. It is strong proof of technical relevance and roadmap access, but it is disclosed as a multi-year development project rather than as a high-volume production account. The unnamed U.S. customer sits somewhere in between: it proves additional demand and geographic expansion, but the lack of identity keeps outsiders from judging concentration, deployment maturity, or whether those orders came from one buyer or several. So the public evidence clears the “real adoption” bar without clearing the “diversified customer base” bar.[CU001, CU002, CU003, CU004, CU005, CU006]

6.2 Adoption trajectory, durability signals, and what remains invisible

The adoption trajectory that can be reconstructed publicly runs from first shipment into Samsung in late 2020, through Korean regional buildout in 2021, through HVM claims by late 2023, into 2024 customer coverage, into a multi-year imec development program in 2025, and finally into 2026 capital for more support and applications capacity. That is a coherent progression. It suggests a company moving from early flagship proof into a broader but still highly technical customer-development phase. What it does not reveal is classic durability data. There is no disclosed customer count, no public NRR or GRR, no renewal schedule, no contract-length data, and no top-customer concentration percentage. The best public durability proxy is repeat-order language: Reuters syndications say Nearfield had multiple orders from an unnamed U.S. customer, and the Series D language emphasizes customer-support buildout rather than one-off research activity. Those are helpful signals, but they are still proxies. There is also no published satisfaction metric, no disclosed service attach rate, and no public account-by-account deployment map that would turn relationship persistence into measurable retention evidence. Public proof is therefore strongest on account quality and weakest on account breadth and persistence. That is acceptable for a technology-diligence story, but not yet sufficient for a concentration-sensitive underwriting model.[CU002, CU007, CU010, CU011, CU012, CU014]

6.3 Expansion motion, support model, and concentration risk

Nearfield’s expansion model appears much more service-heavy than logo-heavy. The company publicly lists offices or representation across major semiconductor geographies, has a visible management bench in sales and operations, and is using fresh capital to build applications centers. Those are all signs of a high-touch customer-support motion suited to large fabs and complex metrology deployments. They are not signs of a broad channel-led model or a low-touch product. In fact, public evidence for channel dependence is relatively thin: Japan and Taiwan references point to commercial agents, but the public record does not show a mature reseller network carrying the load. That leaves concentration risk as the core customer issue. Named proof is high quality but narrow, and the same narrowness that makes the technology feel premium also raises downside sensitivity if a small number of flagship accounts slow down, delay installs, or choose an incumbent workaround. The public evidence therefore supports a constructive but cautious customer verdict. Nearfield has crossed the line from “interesting lab company” to “credible strategic supplier,” yet investors still need the private retention and concentration package before they can judge durability with confidence. Until that package exists, customer strength should be treated as high-quality but concentrated proof rather than as broad-based adoption across many fab programs globally. Breadth remains unproven publicly.[CU009, CU012, CU023, CU024, CU025, CU027]

6.4 Exhibits

7.1 Regulatory and legal risk now sits at the top of the stack

Nearfield’s single most important risk is no longer whether the technology matters. The public record already shows flagship customer proof, repeat orders, and a large 2026 financing round. The bigger issue is whether that commercial momentum can move through a tightening export-control environment without friction. Dutch authorities first broadened the national semiconductor-equipment measure in September 2024 and then tightened it again in January 2025, explicitly saying that specific measuring and inspection equipment would require authorisation from 1 April 2025 for exports outside the EU. Because Nearfield sells advanced metrology and process-control tools and already discloses U.S. and Asian orders, that rule set is directly relevant. The problem is not that the Dutch regime is an outright ban; officials repeatedly say it is case by case. The problem is timing and opacity. Public sources do not disclose Nearfield’s own classification, screening workflow, or approval experience. BIS’s May 2026 guidance adds a second layer by making parent jurisdiction relevant for some advanced-computing customers. That means regulatory drag can arise even when end-market demand remains intact. It also means the diligence burden is operational, not just legal. Someone inside Nearfield has to classify tools, screen end users, monitor parent-company structures, and decide when a promising order is no longer commercially worth the approval effort. None of those controls is visible publicly. For a company selling to a narrow set of technically demanding fabs, that invisibility raises the chance that one delayed transaction matters more than headline funding strength suggests.[CR001, CR002, CR003, CR004, CR005, CR006]

7.2 Order proof is real, but it heightens execution and dependency risk

Operationally, Nearfield looks like a company entering the hard part of scale-up. Repeat orders filled the 2025 order book, a U.S. customer placed multiple system orders, and additional Asian orders were disclosed in 2024. Those are genuine positives, but they also create the next layer of risk: the company must now deliver, install, support, and absorb more geographically distributed demand while still expanding production capacity and applications infrastructure. Public hiring reinforces that reading. The technical project lead role spans manufacturing, procurement, and supply chain, while the optics engineer role runs from concept to volume production. At the same time, the roadmap remains tied to high-value external nodes such as imec and A*STAR IME. Those partnerships are strengths, but they also mean roadmap credibility is partly bound to collaborative programs that the public cannot fully monitor. Nearfield’s support footprint is expanding across the U.S. and Asia, which mitigates response-time risk, yet it also increases management span and knowledge-management complexity.[CR008, CR010, CR011, CR012, CR013, CR014]

7.3 Financing and governance improved, but the underwriting package is still incomplete

The encouraging part of Nearfield’s current profile is that public mitigations are visible. The company raised €135 million in 2024, added a $380 million Series D in 2026, secured an ING facility in between, and added an independent supervisory board member in 2025. Those signals matter because they show that investors, lenders, and partners still want to back the scale-up. The less encouraging part is that these mitigations do not close the central diligence gaps. Investors still do not know the top-customer share, the order-to-revenue conversion cycle, the export-approval timing, or the legal terms inside the bank facility. That means risk monitoring has to stay concrete: watch licence timing, customer concentration, installation cycles, and lender headroom rather than generic “execution” language. In practical terms, Nearfield’s risk profile is high but not thesis-destroying today. Demand proof and strategic relevance are present. What makes the company risky is that a narrow set of external nodes—regulators, flagship customers, lender, and roadmap partners—can still propagate quickly into revenue timing and valuation pressure. The company is therefore investable only if the committee is willing to monitor a handful of measurable tripwires rather than rely on a generic long-duration deep-tech story. If those tripwires stay stable, the risk profile can compress quickly. If they deteriorate, the downside can arrive before public narratives around AI, sovereignty, or unicorn status have time to adjust.[CR020, CR024, CR025, CR026, CR027, CR030]

7.4 Exhibits

8.1 The company story is attractive; the price discipline is the problem

Nearfield is easy to like as a company. The 2026 Series D proved global investor appetite, and company materials keep strengthening the narrative that advanced metrology is becoming strategically central to AI-era semiconductor manufacturing. Repeat orders, a full 2025 order book, lender support from ING, and continued product and whitepaper releases all suggest this is not just a lab story anymore. The challenge is that none of those positives gives public investors the denominator needed to underwrite a $1.6 billion valuation. No revenue, gross margin, cash-burn, or service-mix number was disclosed in the sources reviewed for this run. That means the right investment question is not “Is Nearfield interesting?” It is “What revenue and margin profile must exist for today’s price to be sensible?” Until management answers that question with numbers, the right stance is disciplined curiosity rather than enthusiastic extrapolation. In practice, that means treating Nearfield as a potentially strong business but an only partially priceable security. Strategic quality and entry-price quality are not the same thing yet for outside investors today.[CV001, CV002, CV003, CV004, CV005, CV006]

8.2 Public comps show why the missing denominator matters so much

The comp evidence is directionally useful but not self-executing. Mature process-control and semiconductor-equipment leaders such as KLA and ASML trade at double-digit sales multiples, while Bruker sits much lower and Onto provides a more mid-scale metrology anchor. Because Nearfield’s own revenue is undisclosed, the same $1.6 billion valuation can map to very different implied multiples. At 10x sales, investors are implicitly assuming about $160 million of revenue. At 15x sales, they are assuming a little over $100 million. At Bruker-like instrumentation multiples, they would need more than $750 million of revenue, which is clearly a different business reality. That is why a scenario framework is more honest than false precision. A plausible base case leaves the current price roughly fair-to-stretched. A bull case exists, but it requires stronger revenue conversion and more disclosure. The bear case is not absurd if export timing or order conversion disappoints. Another way to frame the problem is that comps are giving investors a translation key, not an answer sheet. They show how the market prices quality, scale, and disclosure in adjacent businesses, but they cannot tell investors which part of the band Nearfield deserves until management discloses the denominator and margin structure with audited-style detail.[CV007, CV008, CV009, CV010, CV011, CV012]

8.3 Recommendation: research more, and be explicit about what changes the call

The valuation chapter therefore lands on research-more with medium confidence. That is not a disguised “no”; it is a recognition that the company may deserve attention while the security still lacks price support. Nearfield has enough proof to avoid an immediate pass: repeat orders exist, customer demand is not fictional, lender participation suggests some working-capital credibility, and the strategic narrative is coherent. But there is not enough public evidence to justify a buy call at the current round price. The diligence path is straightforward. Revenue, gross margin, order conversion, cap-table preferences, facility covenants, and export workflow all need to be made visible. If management provides those and the numbers support the current multiple, the recommendation can upgrade quickly. If instead repeat orders fail to convert, export approvals drag, or the financing stack proves restrictive, the downside can materialise faster than the headline demand story would suggest. The practical implication is that diligence cannot stop at generic enthusiasm about AI infrastructure. It has to force a hard answer on revenue quality, downside protection, and whether the current price already capitalises most of the visible good news. In other words, the path from research-more to buy is short if the numbers are strong, but the path from research-more to avoid is also short if the hidden economics disappoint. That asymmetry is exactly why the current recommendation remains price-disciplined.[CV018, CV024, CV025, CV026, CV027, CV032]

8.4 Exhibits