The Tri-State has no block above five megawatts left, and it stopped growing.

What the Tri-State is for

Northern New Jersey carries the trading and market-data infrastructure that the financial industry is built on, and Manhattan carries the media, advertising and enterprise demand that wants to sit next to it. Proximity to a named exchange, a named carrier hotel or a named counterparty is the requirement. Nothing about this market is about cheap kilowatts.

Two facts define it right now. CBRE reports that no capacity of more than 5 MW is currently available anywhere in the market, and that Southern California, Austin and San Antonio, and Central Washington have all overtaken the Tri-State on inventory for the first time since CBRE began tracking it in 2016. Under-construction totals were flat against 2024.

Source: CBRE North America Data Center Trends H2 2025 and the CBRE New York Tri-State market profile. CBRE does not publish inventory, vacancy or an asking-rate band for this market in its public releases, so this page carries no figure for those. We would rather leave a gap than fill it with an estimate.

Pricing

CBRE has not published a Tri-State asking-rate band we can cite, so we will not invent one. What it does say is that demand for legacy space is driving record-low vacancy and record-high pricing here, and that 250 to 500 kW requirements in primary markets are expected to exceed $200 per kW per month against a national average of $195.94.

Treat the Tri-State as sitting at or above the top of the national range, and expect the real number to be driven by the building rather than the market. In a carrier hotel with irreplaceable interconnection, the rate reflects what the cross-connects are worth, not what the kilowatt costs.

• Price the cross-connects and carrier ports as a line item. In this market they often exceed the space and power.
• Establish the renewal mechanism before signing. Retention leverage here is almost entirely with the operator.
• Ask what the same requirement costs in a New Jersey suburban campus versus a Manhattan or Newark carrier building, then decide what the milliseconds are worth.
• Confirm the power path, not just the space. Availability of one does not imply the other.

The submarket decision

Northern New Jersey suburban campuses give you more room, better economics and a workable latency position for most workloads that are not doing latency-sensitive trading. Manhattan and the Newark carrier buildings give you interconnection density and the shortest paths, at a rate that reflects it. Orangeburg and the Hudson Valley sit further out with more room to grow.

Recent activity shows the pattern. DataBank has moved to expand its Orangeburg campus after fully leasing the first phase, CoreSite has only a few megawatts left at NY3 with NY4 still under construction, and Equinix is bringing its tenth Tri-State facility to market. Growth is happening, but in increments that get absorbed on arrival.

When the Tri-State is the right answer

When the latency requirement is genuinely specific: a named exchange, a named counterparty, a named carrier hotel, or a user population that will notice. That is a real requirement and there is no substitute for it.

When the driver is that the company is headquartered in New York, it usually is not the right answer. Compare against Northern Virginia for interconnection depth at better economics, and against a suburban New Jersey campus before assuming you need to be in the expensive buildings.

Risks to resolve before shortlisting

• Power first: confirm an energization path before evaluating anything else about a building.
• No room to grow: with no blocks above 5 MW, expansion means a second site. Plan for it now.
• Renewal exposure: a tight market with no alternatives is where renewal terms hurt most.
• Age and density: much of the interconnection-rich estate is old and was not designed for modern density.
• Concentration risk: if your architecture depends on one carrier hotel, understand what a building-level event costs you.

A facility-ready Tri-State brief

1. The latency requirement, stated as a target to a named endpoint.
2. Every cross-connect and the counterparty on the other end.
3. Committed power, and confirmation that it can actually be energized.
4. Density per cabinet against what the building supports.
5. Expansion plan across two sites, since one will not carry it.
6. Renewal and escalator terms, negotiated before the opening rate.

What has been announced in New York Tri-State

Announced capacity is collected by hand from trade reporting and company statements, and each figure links to the report it came from. It is kept apart from the permit record above and never added to it: one is a document a jurisdiction issued, the other is a plan a company published. Neither is availability.

Every New York Tri-State facility we track

97 data centers across New York Tri-State, from the facility database we hold as a Bridgepointe partner. It is the layer that says whether a building exists yet, which no permit record answers directly.

Operators with the most facilities here: Digital Realty (12), Equinix (11), Csquare (7), DataBank (6), QTS (5), 365 Data Centers (4), Cologix (4), CoreSite (4), CyrusOne (4), DataVerge (3).

Where they cluster: New York (16), Secaucus (12), Piscataway (9), Somerset (6), Newark (4), Orangeburg (4).

The largest buildings coming to this market

8 facilities here are under construction or planned. These are the ten largest by floor area, and a size on file is not a promise of a delivery date.

From the Signal facility database, which we hold as a Bridgepointe partner rather than as a public record, so these rows carry no filing link and street addresses are held back. A building existing is not the same as space being available in it: every facility above may be full, pre-leased, or closed to your size. Finding out is a call, not a lookup. See how we verify capacity.