←The mapResearch bank
Theses
Solid-state transformer: unrelieved constraint carrying semiconductor revenue as rent, not as scalePackage Perimeter as Hyperscaler Negotiating Leverage Against NVIDIANVLink's Demand Is a Fragile Coalition of Neoclouds Under Collateral PressureWolfspeed: High-Voltage Moat Evaporates Under Thermal ScrutinySK Hynix: The Memory Supplier Suffering Its Own Customer's ConstraintAccelink: The Valuation Anomaly in China's Optical Chip IndigenizationCoWoS Binds the Buyer, Not the LandlordHBM5: The Transmission-Line Bottleneck Transfers Margin to the Memory SuppliersArm: The Revenue-Mix Pivot Consensus Hasn't PricedInnolight: The Hyperscaler Pricing Ratchet Disguised as GrowthLong-Context Inference: Infrastructure Debt Becomes the LeverageInterconnection queue converts datacenter operator debt into utility equity premiumAmazon: The Capex-Margin InversionAixtron: Customer Capex Discipline Masks Downstream GrowthShin-Etsu: Hidden Exposure to Datacenter Grid FailuresTransformer scarcity is pricing power for industrial conglomerates, not a hyperscaler cost problemElectrical Steel Asymmetry: GE Vernova Captures Transformer Scarcity Without Steel ExposurePower Caps the AI Buildout — the bottleneck moved from chips to megawattsSemicap service is an installed-base annuity, and the market still prices these names as capex cyclicalsThe ABF Chokepoint Does Not Pay Its Owners — Buy the Bottleneck Fails HereABF substrate dual-gate arbitrage: substrate converters capture the spread the market assigns upstreamASML / China-DUV displacement is overpriced — the sell-off prices near-full substitution off a 5-machine data pointNAND Flash: Consumer Demand Collapse Disguises Enterprise Pricing PowerTesla: AI Capex Subsidy Disguised as Automotive TurnaroundUALink: Consortium Activity Masks Structural IrrelevanceCo-packaged optics is a packaging play being misfiled as an optics transitionApplied Optoelectronics: Loss-Funded Market Access Captures Transceiver Supply ScarcityMicrosoft: Free Cash Flow Collapse Telegraphs Margin Compression Before It Hits the P&LRack Power Density: The Liquid Cooling Adoption MirageGPU collateral decay transmits to NVIDIA demand before physical constraints clearHBM moat vs. DRAM commoditization — is SK Hynix's HBM-mix 'miss' actually bullish?Advanced Packaging: The Scarcity Premium Flows Upstream, Not DownSilicon Wafer: Duopoly Ships into Oligopoly Margin Explosion, Captures NoneEaton: Margin expansion telegraphs demand scarcity more than growth can showMoE inference: stranded-power miners own the scarcest input, hyperscalers rent itMurata: The Margin the Bottleneck Hasn't ReachedAlphabet: The Capex Ratchet Liquidity Mispriced as Platform OptionalitySilicon cannot emit light — indium phosphide is the chokepoint CPO relocates but cannot removeEUV scarcity is priced into ASML, invisible in AlphabetOptical attach is set by topology, not by shipments — linear optical TAM models are mis-specifiedConstellation: Hyperscaler Monopsony Masks Nuclear Fuel TransmissionOracle: Monopsony Rent Capture Masked by Consolidated AccountingFoxconn: Customer Concentration Absorbs AI Margin Before It Hits the P&LThe China chokepoint moves from lithography to bonding — and the substitution trade is priced in the wrong categoryNVIDIA's Rent Compresses Through Software, Not SiliconPalantir: The Margin-Protected Infrastructure PlayCustom Silicon Pays a Different MerchantThe Upstream Is Single-Sourced and UnpricedSMIC: Subsidy Converts to Capacity Under Obscured Margin PressureThe Neocloud Rent Is Consumed By The Asset Before It Reaches EquityQualcomm's low capex masks an IP-only future neither consensus nor bears have pricedTransceiver margin expansion is a mirage: buyer concentration at 1.6T resets pricing power the market prices as durableSilicon capacitors expose Intel's package productivity deficitMemory-demand durability — does algorithmic efficiency cap the memory super-cycle?EMIB Shifts the Packaging Bottleneck to SubstratesNeocloud margin re-rating — do open-source deployment + rising rental rates break the thin-margin cap?HBM4 Qualification Concentrates Share, Not Spreads ItCost Per Token Is Set Outside the ChipArista: Hyperscaler capex intensity conceals margin compression riskGlobalFoundries: Customer Concentration Masks Structural Insulation from AI Capex WhiplashThe Megawatt Is the Unit of AccountChipflation Is a Wafer Allocation, Not a Demand Shock
Mechanisms
ABF substrate and build-up film supplyCoWoS advanced-packaging capacityConventional DRAM and NAND supply (HBM crowding-out)EUV tool capacity — the lowest rungGPU residual value as loan collateralGrid interconnection queue positionHeavy-duty gas turbine delivery slotsLarge power transformer lead timesSamsung memory long-term agreementsBuyer concentration tighteningHBF consortiumRent converting into capacityRent migrating upstreamRent not being competed awayUALink ConsortiumUltra Ethernet Consortiumco-packaged-optics displaces copperco-packaged-optics displaces optical-transceiveremib displaces cowosglass-substrate displaces abf-substratehybrid-bonding displaces euvsilicon-capacitor displaces mlccChip designDatacenter mathEfficiency arrives in steps, not trendsInference shapeMemory economicsPhotonicsTau scalingToken mathTokenomicsCapacity arriving — CoWoS (Chip-on-Wafer-on-Substrate)Capacity arriving — HBM4Capacity arriving — Silicon Wafer

EUV tool capacity — the lowest rung

constraint  a named mechanism, not a conclusion

ASML's EUV output is the floor under every gigawatt claim anyone makes. A gigawatt of Rubin-class capacity needs ~2,000,000 EUV passes; one tool does ~591,300 a year at 75 wafers/hour and 90% uptime; so ~3.4 tools per gigawatt. A fleet of ~715 tools by 2030 supports ~211 GW/yr IF EVERY TOOL WENT TO AI. Altman's 52 GW/yr is 25% of that fleet and Musk's 100 GW/yr in space is 47% — two names asking for 72% of every EUV tool on earth.

The path

nodeeffect
Alphabet Inc.-1.00
Datacenter AI (Training + Inference)-1.00
NVIDIA Corporation-0.55
Advanced Micro Devices-0.55
ASML Holding+0.09

What would break it

The scarce thing stops being scarce — capacity arrives, a substitute qualifies, or demand falls to meet supply. Watch the capacity schedule for the same node: a constraint and its schedule are the same fact on two clocks.

coverage and its limits

Scored on 5 subjects: 4 bound by it, 1 own the scarce thing and are made more valuable by it. Severity is graded by hand against a stated scale, not computed — the sign here is a judgement, and the magnitude is that grade.

Where it comes from

Generated from EUV tool capacity — the lowest rung — ASML's EUV output is the floor under every gigawatt claim anyone makes. A gigawatt of Rubin-class capacity needs ~2,000,000 EUV passes; one tool does ~591,300 a year at 75 wafers/hour and 90% uptime; so ~3.4 tools per gigawatt. A fleet of ~715 tools by 2030 supports ~211 GW/yr IF EVERY TOOL WENT TO AI. Altman's 52 GW/yr is 25% of that fleet and Musk's 100 GW/yr in space is 47% — two names asking for 72% of every EUV tool on earth.

The desk already holds CoWoS, power and packaging. Each of those bound in turn and each was relieved, because each is a supply chain that can be started anew: a data centre takes under a year, Amazon has built one in eight months. A fab takes two to three years and the tool that fills it took a decade to make work.

What makes EUV different is not difficulty, it is that the supplier has not been asked to scale and has not chosen to. ASML has never taken price faster than it improved the tool — $150m to $400m while throughput and overlay more than doubled — so there is no margin signal telling the supply chain to expand. The whole chain is running X-1 while the labs are asking for X, and the lag between those is measured in the years it takes to qualify a new supplier of multilayer mirrors.

Two publicly stated ambitions — 52 GW/yr and 100 GW/yr — consume 72% of every EUV tool that will exist in 2030. That leaves 28% for Google, Anthropic, Meta, Amazon, and every phone and PC on earth. The conclusion does not depend on any single input being right; it survives a 25% revision to all of them at once.

Sources

Arguments about this mechanism

None yet. No thesis names this mechanism’s subject, which is a gap in the bank rather than a fact about the mechanism.