The Definitive Guide toAI Data Centers
Ask the GuideAboutAccount

Chapter 2.3

In this chapter · 7 sections
Term help

Long-Lead Procurement & the End-to-End Equipment Supply Chain

In a power-bound build, the longest-lead equipment you have not yet ordered can set the schedule; connect each named vendor and utility milestone to the common energization date instead of applying one portable duration. A factory allocation becomes a schedule input only through a defined order-to-acceptance obligation.

POWER-BOUNDDENSITY-RAMP

What you'll decide here

  1. Which items go on the critical-path lead-time register and can be released before the overall design is complete once their package interfaces are approved—load-serving HV/substation transformers, medium-voltage switchgear, gas turbines and their generator-connected GSUs where generation/export is in scope, and grid-scale chillers/CDUs—versus which can wait for a frozen design.
  2. Whether to commit capital before certainty via reservation deposits and slot-reservation agreements, and how much of the project you are willing to de-risk by paying to hold a manufacturing slot you may not use. State refund rights, draw milestones and what cancels or moves the slot.
  3. Owner-furnished vs contractor-furnished for the long poles: do you take the GPUs, transformers, and gensets onto your own balance sheet and into your own expediting org, or push that risk (and margin) to the EPC? Compare landed cost and the accepted operating envelope on each route.
  4. Single-source vs dual-source vs nearshore for the items where the global supply base is concentrated in two or three factories — transformers, large switchgear, HBM/CoWoS — and what the tariff and export-control overlay does to each option. An available alternative must pass workload and facility-interface acceptance.
  5. Whether you stand up a supply-chain control tower with real expediting, factory acceptance, and heavy-haul logistics, or discover at the gate that your 250-tonne transformer cannot turn onto the site road. Count shared upstream factories, components and ports before crediting a second source.

There is a moment on every AI-campus program when the project manager stops drawing Gantt bars and starts reading a spreadsheet of delivery dates, because the realization has landed that on a power-bound campus the building rarely gates the schedule — the equipment does. Which track actually governs is a property of the dependency network in Chapter 2.1: on a powered-shell fit-out, a brownfield conversion, or a project with an open building permit, it can still be the building. You can pour the slab in eight months, top out the steel in twelve, and pull the fiber in three. None of it matters if the substation power transformer that steps utility HV down to site MV is quoted at ~128 weeks — GSU-class units at ~144–208 — the medium-voltage switchgear at 44, and the gas turbines you planned to bridge with are effectively sold out through 2030. The critical path of a 2026-vintage data center runs through a handful of factories on three continents, and the decision that governs it is when to place the order, against a design that is nowhere near finished.

Long-lead procurement is therefore treated here as a schedule discipline rather than a purchasing one. The pieces: the consolidated lead-time register — the canonical list of what takes how long and therefore what must be ordered first; the contracting instruments that exist precisely because the lead times are intolerable — reservation deposits, slot-reservation agreements, owner-furnished-equipment staging, and buffer inventory; the allocation dynamics that govern the items where demand structurally exceeds supply — GPUs, HBM, CoWoS, transformers, copper; the geographic-concentration and single-source exposures, with the 2025–2026 tariff and export-control regime layered on top; and the control tower that turns a list of POs into delivered, accepted, energized equipment. → grid-side interconnection timing in Chapter 3.2; on-site generation in Chapter 3.4; the contracts that wrap these orders in Chapter 2.4.

The consolidated lead-time register

Every serious owner maintains one artifact above all others in the pre-construction phase: a lead-time register that lists each major equipment package, its current quoted delivery, the design maturity required to release the order, and the latest date you can place it without slipping energization. The live, dated version of that register — every package below, resolved to its source claim and wrapped in a reverse-scheduling planner that turns a ready-for-service date into the PO dates it implies — is the lead-times page. It is the document from which the master schedule is reverse-engineered. The hard truth it encodes is that the long poles must be ordered before the design that specifies them is complete — you commit to a transformer's MVA rating, impedance, and winding configuration while the single-line diagram is still at 60%, because if you wait for 100% the slot is gone.

The register sorts naturally into tiers. The grid-tie tier — HV and generator step-up (GSU) transformers, high-voltage switchgear, and the utility-side substation equipment — sits at the top, with the longest and most volatile lead times of anything in the building. The power-chain tier — distribution transformers, medium-voltage switchgear, UPS, and busway — is shorter but still measured in many months. The on-site generation tier — gas turbines and reciprocating engines for bridge or primary power, plus diesel gensets for backup — has become a critical path in its own right since the gas rush of 2024–2025. The cooling tier — chillers, cooling distribution units (CDUs), and dry coolers — gates the density the building can actually support. And the compute tier — GPUs, and upstream of them HBM and CoWoS advanced-packaging capacity — is governed less by manufacturing time than by allocation.

The consolidated long-lead register (2026 practitioner quotes)
Equipment packageTypical lead time (2026)Design maturity to orderWhy it is the long pole
HV power transformer~128 weeks (≈2.5 yr); GSU ~144–208 wk (SemiAnalysis mid-2026: 3–4 yr US); up to ~60 mo constrained~60% single-line; ratings frozenGlobal grain-oriented-steel + bushing bottleneck; 2–3 dominant OEMs
HV / grid-scale switchgearHV breakers ~100–155 wk (little improvement since the ~151-wk late-2023 peak); HV GIS / full substations 3–5 yr in constrained markets; MV ~44–65 wk~60–70% electrical designBreaker and relay supply concentrated; HV is the bottleneck, not MV
Gas turbine (frame, 290–430 MW)Delivery 2028–2030; OEMs sold out toward 2030Site + emissions basis; turnkey scopeGE Vernova 116 GW backlog + slot reservations at Q2 2026, ≥125 GW targeted by year-end; slots reserved, not bought
Aeroderivative turbine (25–57 MW, LM2500/LM6000)18–36 mo+; refurb under 12 mo but scarceBridge-power basisThe fast on-site path; WoodMac quoted some turbine classes at ~243 weeks (Q2 2025)
Diesel genset (1.25–3.25 MW)52–70 weeksBackup topology + tierTier-4 engine + alternator supply; major OEMs booked out 1–2 yr with capacity expansions underway
Grid-scale chiller / CDUChiller ~20–60 wk (500+ ton centrifugal/screw at the top of the band); CDU ~16–52 wk and rising with liquid-cooling demandDensity tier + cooling modality frozenCaps the density the hall can host; CDU is new-and-allocated
GPU rack-systems (NVL72-class)~6–9 mo from an allocated order to a deployed multi-rack program; a single accepted GB300 NVL72 order against a confirmed slot runs ~12–24 wk to dock (see the lead-times tracker)Cluster sizing + power/cooling envelopeGated upstream by HBM/CoWoS allocation, not assembly time
HBM / CoWoS (upstream of GPUs)GB200/GB300 use HBM3E and Vera Rubin uses HBM4; secure memory and CoWoS capacity before fixing the rack dateN/A — secured via allocation, not POContract memory and packaging capacity before treating the rack date as firm
Lead times are 2025-Q2 to 2026 market quotes and vary enormously by spec, region, and whether you hold a reservation. Sources: Wood Mackenzie T&D survey; GE Vernova / Rolls-Royce investor disclosures; JLL equipment lead-time index; Micron, SK hynix, and TSMC supplier disclosures. Cross-check each against a live quote — these move quarterly.

Procurement as a schedule discipline

Once you accept that the equipment gates the schedule, three contracting behaviors follow that look reckless to a traditional procurement organization but are the rational response to the lead times. Each one trades capital-at-risk for schedule certainty — and the fork is how much of that trade you are willing to make.

Reservation deposits. The first behavior is paying a non-refundable or partially-refundable deposit to hold a manufacturing slot before the design is frozen and sometimes before financing closes. For a transformer or a switchgear lineup this might be 10–30% of the unit price; for a gas turbine it can be structured as a milestone toward a binding order. The deposit buys you a place in the queue. The risk is obvious: if the project slips or dies, you may forfeit the deposit or own equipment you cannot use. Not paying it has an equally concrete outcome — you lose the slot and re-enter the queue at the back, adding one to three years to energization.

Slot-reservation agreements as a distinct instrument. The most important contracting development of the 2024–2026 turbine rush is that the slot reservation has become its own legal instrument, separate from the equipment supply agreement. GE Vernova ended Q2 2026 with 116 GW of combined gas-turbine backlog and slot-reservation agreements and raised its year-end target to at least 125 GW — effectively sold out through 2030, with residual open slots now quoted into 2031. Much of that book is reserved capacity, not yet converted to firm turnkey orders. Siemens Energy printed the same pattern days later: a 69 GW firm gas-turbine backlog (15 GW booked, 6 GW shipped in the quarter), lead times of "three years or more," and a €51B Grid Technologies book with transformers the largest order contributor (Q3 FY2026, reported 2026-08-05). A slot-reservation agreement lets a developer lock a delivery window with a deposit and a schedule of conditions, then convert it to a full supply contract once the site, permits, and offtake mature. It is a financial option on manufacturing capacity. Caterpillar–Hunt Energy and Rolls-Royce's 2027–2028 order book show the same pattern in engines. → the contract mechanics in Chapter 2.4.

OFE staging and buffer inventory. The third behavior is decoupling the order date from the install date by taking delivery early and staging equipment — owner-furnished-equipment (OFE) staging — and by holding buffer inventory of the items that fail or get damaged. You order the long-lead transformer to arrive when the pad is ready, not when you finally need it energized, and you carry critical spares (a spare transformer for a campus, CDU pumps, switchgear breakers) because a single failure of an item with a two-year lead time is an existential schedule risk, not a maintenance event. The cost is carrying inventory and storage; the alternative is a two-year hole in the schedule when a unit arrives damaged.

~144–208 wk
GSU transformer lead time (standard power transformer ~128 wk); up to ~60 mo in constrained markets
the schedule's long pole — order it against a 60% design or accept the slip
Scope & caveats

GSU units only, where on-site generation or export is in scope. The load-serving substation transformer is the separate claim hv-substation-power-transformer-lead-time-144 (~128 wk average, ~128–208 wk).

Indices diverge in mid-2026: VAWN's August index still lists GSUs at 144 wk (power/substation 160+ wk) while SemiAnalysis reports US GSU lead times of three to four years. Treat the bottom of the range as reservation-holding buyers and the top as new orders.

~100–155 wk
HV switchgear & circuit-breaker lead time — HV breakers with little improvement since the ~151-wk late-2023 peak; HV GIS and full substations 3–5 yr
the second pole, and the one most often planned at MV lead times by mistake
~33–42 weeks
JLL global / US average data-center equipment lead time across packages
the baseline every package is measured against — anything above it is a critical-path candidate
52–70 weeks
diesel genset lead time (1.25–3.25 MW class); largest engines to ~78 wk
Tier-4 engine and alternator supply is heavily booked — backup power is a long-lead package too
~1M wafers
estimated 2026 CoWoS demand — ~fully booked for the year; NVIDIA holds ~50–60% of allocation
allocation, not assembly, sets your delivery date
~16–52 wk
CDU lead time (direct-to-chip liquid cooling) — 16–24 wk standard, 26–52 wk in the current queue
a new-and-allocated category: the liquid-cooling plant can gate a hall the power chain is ready for
~12 mo
Blackwell (B200/GB200) allocation backlog for new orders at launch — the historical allocation case, not a current quote; price a 2026 decision from the live lead-times tracker
GPUs arrive on an allocation, not a quoted PO — the date is negotiated upstream, not scheduled
Scope & caveats

October-2024 launch-period backlog for new B200/GB200 orders. A historical allocation datapoint, not a current quote: schedule a 2026 procurement decision from a product-specific, clock-defined entry on the live lead-time register.

sold out
NVIDIA cloud GPU availability, all shipping generations — 'Blackwell sales are off the charts, and cloud GPUs are sold out'
a large contiguous cluster of the generation and fabric you need cannot be assumed available on demand — evaluate a cloud bridge as its own decision, not as a fallback
HBM3E sold out
Micron pricing agreements covering the vast majority of its calendar-2026 HBM3E supply (September 2025)
supplier allocation and pricing agreements set the order window
116 GW (Q2 2026)
GE Vernova gas-turbine equipment backlog + slot reservations (116 GW combined at Q2 2026)
on-site generation is sold out toward 2030 — a bridge you must have booked already
Scope & caveats

Combined equipment backlog and slot-reservation agreements; GE's Q2 release does not publish a firm-versus-reserved component split.

50% tariff
US Section 232 duty on semi-finished copper (Aug 2025); electrical grid equipment got a 15% transitional rate
a policy shock that lands on the busway and cable you already priced

Allocation dynamics and vendor financing

For a transformer or turbine, the quoted N weeks combines queue position, approved drawings, component supply, production, factory testing and logistics; it is not all manufacturing work. For a compute package without a secured factory slot, the governing variable is allocation. When the vendor rations a sold-out GPU configuration, willingness to pay does not establish delivery; the buyer needs an allocation for the specified SKU and rack system. The real gate sits upstream of the accelerator itself, at HBM (high-bandwidth memory) and CoWoS (TSMC's advanced 2.5D packaging that bonds the logic die to the memory stacks). GB200 and GB300 use HBM3E, while Vera Rubin uses HBM4; Micron said in September 2025 that it had pricing agreements with almost all customers for the vast majority of its calendar-2026 HBM3E supply, while TSMC's 2026 disclosures describe continued CoWoS capacity expansion. The accelerator’s quoted ~6–9-month order-to-delivery window must identify the allocation and starting event before it can enter the IMS. → the silicon and memory supply detail in Chapter 7.6; advanced-packaging engineering in Chapter 7.7.

Allocation creates a power dynamic that bleeds into the rest of the deal. Large buyers secure GPUs through long-term commitments, prepayments, and increasingly vendor financing — the chip vendor or a partner extends credit, takes equity, or co-invests in the buyer so the buyer can afford the order that the vendor will then fulfill. This circularity (the supplier financing its own demand) is both a genuine mechanism for clearing supply and a structural risk the finance chapters scrutinize closely. For the procurement organization, the practical consequence is that the GPU order is a relationship and a capital commitment, negotiated quarters ahead, and the buyers who are inside the allocation are the ones who committed capital and volume early. → the circular-financing debate and GPU-cloud demand underwriting in Chapter 2.5; the residual-value and depreciation stress in Chapter 1.8. A cloud bridge must separately qualify contiguous capacity, topology, data transfer, workload acceptance, price and exit date. Renting an available accelerator recovers schedule only if the job can run on the offered system.

Deep dive: why CoWoS and HBM — not the GPU — set your delivery date

It is tempting to think of GPU supply as an assembly problem: TSMC fabricates the logic die, the memory makers stack the HBM, an OEM bolts seventy-two of them into an NVL72, and you receive racks in 6–9 months. Do not assume any step has slack: CoWoS and HBM are two upstream constraints to test against the supplier’s complete delivery commitment.

The first is CoWoS — chip-on-wafer-on-substrate, the 2.5D packaging step that places the GPU logic die and the HBM stacks side-by-side on a silicon interposer. NVIDIA's current rack-scale accelerators depend on it, and TSMC continues to expand CoWoS capacity against strong AI demand. The second is HBM itself — the stacked DRAM that gives the accelerator its memory bandwidth. Micron said in September 2025 that it had pricing agreements with almost all customers for the vast majority of its calendar-2026 HBM3E supply; SK hynix identifies HBM3E with GB300 and HBM4 with Vera Rubin.

The consequence for the procurement organization: TSMC’s CoWoS allocation and the HBM makers’ order books can constrain the GPU delivery date quarters before shipment; OEM integration, testing and transport still have to close. If the 6–9-month assembly quote excludes an unsecured allocation, it is not a complete delivery commitment — the real clock started when someone secured packaging and memory capacity, and that someone was probably the hyperscaler ahead of you in line. → upstream silicon in Chapter 7.6; packaging in Chapter 7.7. Require the allocated SKU, component dependencies and firm factory slot in the procurement record. A substitute that changes rack power, residual-air load, coolant conditions, optics or firmware reopens the relevant acceptance baseline; the same accelerator count does not prove interchangeability.

Geographic concentration & single-source risk

The lead times above are bad because the supply base is thin and concentrated. Several of the most critical packages come from a handful of factories, often in a handful of countries, and a disruption at a shared source can move several supposedly independent project dates together. This is the structural risk that sits underneath the schedule risk, and it forces a sourcing fork on the items where it bites hardest.

Transformers are the canonical example: a small number of OEMs, a global shortage of grain-oriented electrical steel, and bushing and tap-changer sub-suppliers that are themselves single points of failure. A fire, a labor action, or an export restriction at one shared plant can move delivery dates across several regions. Grid-scale switchgear and HV breakers are similarly concentrated, which is why full HV substations and GIS lineups in constrained markets quote up to 3–5 years (individual HV breakers typically ~100–155 weeks) while medium-voltage gear has eased. The copper crunch sits underneath both: an NVL72 rack alone carries thousands of in-rack copper NVLink cables, the power chain is copper-intensive end to end, and copper is simultaneously demanded by every electrification and grid-replacement project on the planet — a demand collision that the 2025 tariff regime then taxed. And the HBM/CoWoS gate concentrates the entire compute tier into a few fabs in Taiwan and a few memory plants in Korea and the US. Count independent failure paths, not bidder logos: record the awarded factory, critical upstream source, shipping port and qualified recovery option. A second assembler using the same transformer core or HBM supply does not separate that failure.

Sourcing strategy by concentration risk
ItemConcentrationSingle-source riskPractical mitigation
HV / GSU transformerQualified OEM count depends on voltage, rating and utility approval; GO-steel + bushing bottleneckA plant disruption moves dates worldwide; no fast substituteReserve slots early; carry a campus spare; qualify a second OEM
HV / grid-scale switchgearFew breaker/relay makers; HV the bottleneckup to 3–5 yr quotes for HV GIS; relays a sub-tier SPOFStandardize on a dual-OEM spec; pre-order long-lead breakers
Copper (cabling + power chain)Global commodity, now tariff-taxed; smelting concentratedPrice + availability shock; Section 232 50% dutyForward-buy / hedge; design for content efficiency; nearshore mills
HBM / CoWoS / GPUTSMC packaging + 3 HBM makers; one vendor ~50% of CoWoSWithout a confirmed HBM/CoWoS and OEM allocation, a higher chip price alone does not establish a delivery dateMulti-vendor accelerator strategy; long-term volume commitments
Gas / aeroderivative turbineHandful of OEMs; sold out toward 2030Slot is the asset; latecomers wait yearsSlot-reservation agreements; refurb/aero bridge units
The fork is single-source (cheapest, fastest to contract, highest risk) vs dual-source vs nearshore. The right answer differs by item because the supply base differs.

Tariffs, trade policy & export controls

On top of thin supply and long lead times, 2025–2026 layered a volatile trade-policy overlay that changes the price and sometimes the legality of imported equipment. The procurement organization can no longer treat a quoted price as the price — it must model the tariff and export-control regime that applies on the day the goods clear customs, which may differ from the day the order was placed.

The most material change for electrical equipment was the US Section 232 action on copper: from August 2025, a 50% tariff on semi-finished copper products and copper-intensive derivatives (cables, connectors), recalibrated in April 2026 to apply to the full value of semi-finished products. For builders, electrical grid equipment received a 15% transitional rate through end-2027 to avoid kneecapping the very buildout the policy was meant to protect — a carve-out worth knowing, because it changes the math on imported switchgear and transformers versus domestic. Steel and aluminum Section 232 duties stack on top of the structural-steel and enclosure costs. On the compute side, export controls on advanced accelerators and the equipment to make them reshape which chips can ship where — a procurement-defining constraint for any operator serving, or sourcing through, a controlled jurisdiction.

The strategic responses are nearshoring and dual-sourcing: qualifying a domestic or allied-country transformer and switchgear OEM to escape both the tariff and the single-source exposure, and accepting a higher unit price or a different lead-time profile in exchange for tariff certainty and supply resilience. The tradeoff is the familiar one — the cheapest imported unit may carry a tariff and a concentration risk that the slightly-more-expensive nearshore unit does not, and the right answer depends on the tariff trajectory you believe and the schedule you can tolerate. → permitting and the regulatory critical path in Chapter 3.9, and export controls and sovereignty in Chapter 3.12; the contractual allocation of tariff and change-in-law risk in Chapter 2.4.

Owner-furnished vs contractor-furnished for the long poles

Who actually buys the long-lead equipment — the owner, or the EPC contractor — is a decision with direct schedule and risk consequences, and it is decided package by package, not once for the whole project. Owner-furnished equipment (OFE) means the owner places the order, takes title, carries the deposit and the inventory risk, and hands the equipment to the contractor to install. Contractor-furnished pushes the procurement, the float, and the supplier-management burden onto the EPC, who prices that risk into the contract.

The logic divides cleanly by item. GPUs are almost always OFE — they are the most valuable, most allocation-constrained, most rapidly-depreciating asset in the building, and the owner who controls the allocation relationship is not going to delegate it. Transformers and gas turbines are frequently OFE on AI campuses precisely because the owner wants to reserve the slot and place the order long before an EPC is even selected — waiting for the contractor to procure the long pole would forfeit the schedule. Chillers and CDUs are split: owners increasingly furnish them to lock density-critical capacity early, but they sit closer to the contractor's mechanical scope. The price of OFE is that the owner now owns the expediting, the factory acceptance, the logistics, and — critically — the interface warranty gap: when an owner-furnished transformer fails to integrate with a contractor-installed switchgear lineup, the finger-pointing between supplier and installer lands on the owner. That interface-warranty exposure is the hidden cost of OFE, and it is contracted around explicitly. → OFE vs contractor-furnished in the delivery-model framing of Chapter 2.2; the interface-warranty mechanics in Chapter 2.4.

The supply-chain control tower

A purchase order is a promise, not a delivery. The gap between the two — manufacturing slippage, failed factory acceptance, a transformer that cannot physically reach the site — is closed by a supply-chain control tower: a dedicated function that tracks every long-lead package from order to energization and intervenes when reality drifts from the plan. On a single-building project this might be one expeditor; on a multi-gigawatt campus it is a standing organization with three core jobs.

  • Expediting. Active, on-the-ground tracking of manufacturing progress against milestones — not waiting for the supplier's status email, but visiting the factory, verifying that the GO-steel arrived and the windings are on schedule, and escalating the moment a sub-supplier slips. For an item with a two-year lead time, discovering a six-week slip in month four is recoverable; discovering it at the promised ship date is a disaster.
  • Factory acceptance testing (FAT). Witnessing the equipment pass its acceptance tests at the factory before it ships — a transformer's impulse and heat-run tests, switchgear functional checks, a CDU's flow and pressure verification. A unit that fails FAT and goes back into the queue is a schedule event measured in months; catching it at the factory is far cheaper than catching it at the gate. → FAT's place on the commissioning ladder in Chapter 13.1.
  • Logistics of oversized and heavy loads. The control-tower job programs most often overlook. A large power transformer can weigh 200–400 tonnes; a fully built NVL72-class rack ships at roughly 1.4 tonnes. That mass triggers route and structural verification, but acceptance depends on the OEM foot/wheel and rolling reactions, temporary rigging states, and the complete floor assembly—not on calling the cabinet mass one point load. These are super-loads requiring route surveys, permits, escorts, sometimes bridge reinforcement, and a confirmation — done early — that the equipment can physically turn onto the site road and through the building opening. Programs have discovered, at the gate, that the transformer cannot make the final turn. The control tower verifies the route before the order ships, not after.
The grid-tie equipment timed here is part of the interconnection critical path detailed in Chapter 3.2, and the on-site turbines and gensets are the speed-to-power bridge engineered in Chapter 3.4. The substation and power-chain equipment is specified in Chapter 4.2; the cooling plant and CDUs in the cooling chapters. The contracting instruments — slot-reservation agreements, equipment supply agreements, OFE interface warranties, and change-in-law / tariff risk allocation — are the subject of Chapter 2.4, and the capital that funds the deposits and prepayments is structured in Chapter 2.5. The GPU and HBM allocation game runs upstream into silicon in Chapter 7.6 and advanced packaging in Chapter 7.7; the OFE-vs-contractor-furnished delivery framing is set in Chapter 2.2; and the depreciation and residual-value stress that makes every month of schedule slip expensive is quantified in Chapter 1.8.
Cite this chapter
Fehn, J. (2026). Long-Lead Procurement & the End-to-End Equipment Supply Chain (Chapter 2.3). The Definitive Guide to AI Data Centers. https://aidatacenterguide.com/part-2-project-delivery-schedule-procurement-contracts-and-risk/2-3-long-lead-procurement-and-the-end-to-end-equipment-supply-chain (accessed 2026-09-29).
@misc{aidc-2-3,
  author       = {Fehn, Jacob},
  title        = {Long-Lead Procurement & the End-to-End Equipment Supply Chain (Chapter 2.3)},
  howpublished = {The Definitive Guide to AI Data Centers},
  year         = {2026},
  url          = {https://aidatacenterguide.com/part-2-project-delivery-schedule-procurement-contracts-and-risk/2-3-long-lead-procurement-and-the-end-to-end-equipment-supply-chain},
  note         = {Accessed 2026-09-29}
}
Spotted an error? Suggest an edit