NTT Mumbai 5 in Chandivali is 12.5 MW of critical IT load, operational, and it counts in full. It is one of eight NTT Mumbai records in this registry that do, totalling 161 MW — every one of them IT load stated, stage stated and dated. In the same city, STT publishes 133 MW and counts zero, because every STT record sits inside a portfolio aggregate. Two operators, one market, and a 161 to 133 difference in published capacity that becomes a 161 to nothing difference in the registry. NTT is what the disclosure standard looks like when it is actually followed.
1At a glance
2Two ways to publish the same city
This is the comparison the whole registry has been building toward, and it happens entirely inside Mumbai.
| NTT Mumbai | STT GDC Mumbai | |
|---|---|---|
| Published capacity | 161 MW | 133 MW |
| Records in this registry | 8 | 6 |
| Basis stated | critical IT load, every record | power capacity, unlabelled |
| Stage stated | operational, every record | stage unverified, every record |
| Portfolio aggregate flag | none — each building stands alone | non-additive, and every child too |
| Counted in the state total | 161 MW | 0 MW |
The capacity figures are close: 161 against 133. Everything else is opposite. NTT publishes eight buildings and lets each one stand. STT publishes a portfolio and files every building underneath it.
The registry does exactly what it was told to do. It counts facilities and refuses to count aggregates. NTT looks like a 161 MW operator and STT looks like it has no Mumbai presence at all.
3Why NTT is countable
Nothing NTT does is heroic. It publishes four facts per building, which is the minimum for an asset to be usable as data.
| Fact | NTT Mumbai | CtrlS Mumbai | STT Mumbai |
|---|---|---|---|
| Capacity per building | yes | yes | yes |
| Basis stated (IT or facility) | IT, explicit | unlabelled | unlabelled |
| Stage stated | operational | unverified | unverified |
| Facility vs facility power separated | yes where both exist | no | no |
| Portfolio aggregate that swallows it | none | the 700 MW campus claim | the city headline |
The fourth and fifth rows are what make the difference. NTT separates IT load from total facility load where both are known — Mumbai 7 is 25.6 MW of IT in a 38 MW facility — which is the single most useful disclosure an operator can make, because it gives you the PUE.
And NTT does not publish a city portfolio that its buildings sit inside. There is no 161 MW headline for Mumbai. There are eight buildings, and a reader adds them up if they want to.
Four facts per building, no aggregate overhead. That is all it takes to be countable.
Two operators in this file do it: NTT and, for one building, STT Jaipur. Everyone else publishes a portfolio and disappears from the totals.
4The eight NTT Mumbai buildings
Mumbai 5–9 plus three small exchange sites make up the NTT Mumbai count. This is what a properly itemised metro portfolio looks like.
| Record | MW | Basis | Location |
|---|---|---|---|
| Mumbai 5 | 12.5 | critical IT load | Chandivali |
| Mumbai 6 | 19.2 | critical IT load | Andheri East |
| Mumbai 7 | 25.6 | critical IT load in a 38 MW facility | Chandivali |
| Mumbai 8 | 36 | facility capacity, 2,800 racks | Chandivali |
| Mumbai 9 | 62 | facility capacity, 6,000 racks | Chandivali |
| BSE Data Center | 0.3 | critical IT load, 1 MW facility | Fort |
| Vikhroli | 2.4 | critical IT load, 6 MVA facility | Vikhroli |
| Vikhroli IV | 3 | critical IT load, 4 MVA facility | Vikhroli |
| Total counted | 161 | — | — |
Note the basis column. Three different bases appear — critical IT load, facility capacity, and a mix — and each is stated. That is not perfect consistency, but it is honest: a reader knows which figure they have and can decide whether to compare it.
It is also why Mumbai 8 and 9 cannot be compared directly with Mumbai 5, 6 and 7. They are facility figures, and at a typical PUE they represent materially less IT load. The distinction is published rather than hidden, which is the whole point.
5What this proves
NTT is the existence proof that the disclosure problem is a choice rather than a constraint.
| Objection | NTT’s position |
|---|---|
| Publishing capacity per building is commercially sensitive | eight buildings published, per building, all operational |
| IT load and facility load are hard to separate | both published where both are known, including 25.6 in 38 |
| Customers do not want their facility identifiable | the capacity is published; the customers are not |
| A portfolio headline is simpler for buyers | there is no portfolio headline — the buildings are the message |
None of what NTT publishes compromises a customer or reveals a price. It reveals what the asset is, which is what anyone buying, financing or regulating a data centre needs to know.
The commercial consequence is direct: NTT appears in every state total, every market share calculation and every density comparison this registry produces. STT appears in none of them. A buyer using this file would conclude NTT is the larger Mumbai operator, and on published capacity it is not by much — 161 against 133.
6Procurement
7Limits and open questions
- No rack count or floor area for this building, though other NTT records publish racks.
- No PUE stated, though it is derivable for Mumbai 7 from the IT and facility pair.
- No commissioning date on the card for Mumbai 5.
- Basis varies across the eight records — IT for some, facility capacity for others.
- No customer information, correctly.
- What would close it. A rack count and a PUE. NTT publishes both for other sites, so this is a gap rather than a policy.
8Sources
Sources: NTT Global Data Centers Mumbai facility pages, NTT India corporate site, Data Center Dynamics reporting on the Mumbai 7 and Mumbai 8 launches, and Indian market commentary on the Mumbai data centre market.