Four terms decide almost everything about how reliable a facility has to be, and therefore how much engineering difficulty — and margin — sits behind the equipment that keeps it running.
The Uptime Institute's Tier classification (Tier I through Tier IV) and the telecom-industry standard TIA-942 both certify the same underlying thing from slightly different angles: how much redundant, concurrently-maintainable power and cooling capacity a facility has, not how large it is. A Tier III facility can take any single piece of power or cooling equipment offline for maintenance without an outage; Tier IV adds fault tolerance against an unplanned equipment failure on top of that. Every rung up this ladder means more transformers, more switchgear, more chiller capacity and more cable pathways than the facility's raw IT load would otherwise require — reliability is engineered redundancy, and redundancy is exactly the demand this report's companies are selling into.
N+1 means one spare unit beyond what peak load requires (one extra chiller, one extra UPS module); 2N means a fully duplicated, independent second system that can carry the entire load alone. A hyperscale campus targeting Tier III/IV reliability typically specifies 2N on its most critical power path and N+1 further out — and every step up this specification is a direct, engineered order for more transformers, more switchgear and more cooling capacity than a simpler facility would ever need.
A transformer's capacity is quoted in MVA (megavolt-amperes); GIS (gas-insulated switchgear) is the compact, sulfur-hexafluoride-insulated alternative to open-air switchgear, used where space is constrained or reliability requirements are highest. Reading a company's order book in MVA and GIS-bay terms, rather than in generic rupee figures, is the closest this industry gets to reading a chip company's wafer-capacity numbers — it is the physical unit the scarcity in this report actually lives in.
Power Usage Effectiveness (PUE) is the ratio of total facility power draw to the power actually delivered to IT equipment; a PUE of 1.5 means every unit of computing draws an additional half-unit of power for cooling, power conversion losses and other overhead. Every fraction of a point a cooling or power-distribution company can shave off a hyperscale campus's PUE is worth real, recurring money to that campus's operator over the facility's lifetime — which is exactly the commercial argument the cooling and power-equipment companies in this report make to their own customers.
| Term | What it actually certifies | Who specifies it | What it does NOT tell you |
|---|---|---|---|
| Tier III/IV, TIA-942 | Redundant, concurrently-maintainable (or fault-tolerant) power and cooling capacity | The facility operator/developer, verified by Uptime Institute or a TIA-942 auditor | Whether the equipment inside actually came from an India-listed supplier, or was imported |
| N+1 / 2N | How much duplicate capacity sits behind the critical power and cooling path | The facility's own engineering specification, set at design stage | Whether that duplicate capacity was sourced competitively (commodity) or from a scarce, qualified supplier (engineered) |
| MVA / GIS | The actual physical capacity and switching technology installed | The power-equipment supplier's own nameplate rating | Whether current order backlogs reflect genuine scarcity or ordinary cyclical demand — see §6-7 |
| PUE | How much of a facility's total power draw is computing versus overhead | Independent, ongoing measurement by the facility operator | Which specific vendor's equipment is responsible for a given PUE improvement — attribution is rarely disclosed |
Source: Dart Consultants, from Uptime Institute and TIA-942 published standard summaries, and industry-standard PUE/MVA/GIS definitions as used across this report's company research — see Notes for full citation list.
If a company's revenue comes from meeting a specification someone else wrote — a building shell, a cable run, a civil contract bid competitively against several other qualified contractors — assume the margin is thin and under continuous pressure, no matter how large the order book looks. If a company's revenue comes from equipment whose global manufacturing capacity is itself constrained — a transformer, a GIS bay, a purpose-engineered cooling system — assume that is where this report's attention should go, and where real, defensible pricing power is more likely to actually show up in the numbers.
