Auto-generated from live data. 21,417 facilities tracked, 233 markets scored by the DC Hub Power Index (DCPI), 144 M&A deals, 4 hyperscaler $1B+ deals in this quarter. Generated 2026-05-30 15:20:53 UTC.
The dominant shift in Q2 2026 is a decisive capital concentration play: $181.2 billion in M&A activity across 144 transactions, with hyperscalers and mega-PE firms claiming outsized share. Google's $40 billion acquisition and Blackstone's combined $41 billion in two separate deals set the tone—this is not passive portfolio management but aggressive competitive positioning for power-constrained capacity. Against this capital intensity, new facility construction remains anemic: zero new facilities entered tracking this quarter, while the announced and planned pipelines total just 16 gigawatts of incremental supply. The market verdict is stark: 150 markets rated CAUTION, 75 rated AVOID, and only 8 rated BUILD. Capital is flowing, but deployment discipline is tightening. Operators are consolidating, not expanding recklessly.
This window exposes a structural realignment: the power gap is now the binding constraint, and geography has become destiny. The five BUILD markets—Montreal, Overland Park, Tulsa, Lenexa, and Oklahoma City—cluster in two regions with superior power economics: Canadian hydro (HQ) and SPP's wind-rich, lower-cost footprint. Conversely, Western Europe's entire top five (Dublin, London, Amsterdam, Frankfurt, Manchester) sits in the AVOID column, signaling that premium tier-one markets can no longer command capital allocation absent solved power infrastructure. The verdict distribution confirms it: 64% of scored markets warrant caution or avoidance, not opportunism. Hyperscalers are chasing power availability and regulatory certainty, not density. This tilts the next four quarters toward build-out in secondary, power-advantaged regions and selective infill in constrained tier-one markets only where power PPAs exist or are imminent.
Monitor SPP market absorption rates closely through Q3 2026—the concentration of four BUILD verdicts in Kansas and Oklahoma is not coincidental and will test whether these markets can absorb the capital flowing from coastal congestion. Watch for the first major hyperscaler or PE buyer to announce a multi-GW, multi-year commitment in that region; such an anchor would validate or break the emerging thesis. Second, track European M&A stalls: if the AVOID verdict holds across London and Amsterdam through year-end despite pricing compression, it signals permanent structural damage to those hubs and will force European operators into distressed asset sales. Third, monitor the Blackstone and Ares playbooks—they deployed $49.6 billion combined; if their next deployment targets power-constrained markets anyway, the power-first thesis breaks. If they follow capital into SPP and secondary regions, the market has answered its own question about where profit lies.
Every market gets scored daily. BUILD markets pass excess-power and time-to-power thresholds. CAUTION = mixed signals. AVOID = grid-constrained or capacity-saturated.
Ranked by composite DCPI score. Excess Power + Constraint + TTP (time-to-power) sub-scores shown.
| Market | ISO | Composite | Excess Pwr | Constraint | TTP |
|---|---|---|---|---|---|
| Montréal, QC | HQ | 39.6 | 65 | 26 | 8 |
| Overland Park, KS | SPP | 25.5 | 58 | 33 | 14 |
| Tulsa, OK | SPP | 25.5 | 58 | 33 | 14 |
| Lenexa, KS | SPP | 25.5 | 58 | 33 | 14 |
| Oklahoma City, OK | SPP | 25.5 | 58 | 33 | 14 |
| La Vista, NE | SPP | 25.5 | 58 | 33 | 14 |
| Papillion, NE | SPP | 25.5 | 58 | 33 | 14 |
| Omaha, NE | SPP | 25.5 | 58 | 33 | 14 |
Markets where DCPI flags grid constraints, capacity saturation, or interconnect queue delays.
| Market | ISO | Composite | Constraint |
|---|---|---|---|
| Dublin, OH | PJM | -50.1 | 72 |
| London, UK | NGESO | -58.3 | 72 |
| Amsterdam, NL | ENTSOE-NL | -56.3 | 70 |
| Frankfurt, DE | ENTSOE-DE | -47.4 | 67 |
| Manchester, NH | ISONE | -48.0 | 66 |
| Phoenix, AZ | WECC | -26.7 | 62 |
| Sydney, AU | AEMO | -28.1 | 61 |
| Tokyo, JP | TEPCO | -48.7 | 60 |
| Singapore, SG | EMA | -53.7 | 59 |
| Chicago, IL | PJM | -38.4 | 56 |
Raw facility counts and aggregate MW grouped by reported lifecycle status.
| Status | Facility Count | Aggregate MW |
|---|---|---|
| operational | 11,063 | 91,116 MW |
| active | 10,055 | — |
| under construction | 154 | 33,676 MW |
| planned | 91 | 27,896 MW |
| announced | 50 | 14,339 MW |
| under development | 2 | 2,200 MW |
| expanding | 1 | — |
| approved | 1 | 1,600 MW |
Concentration leaders. M&A activity in this group drives most market-share shifts.
| Operator | Facilities | Aggregate MW |
|---|---|---|
| Digital Realty | 674 | 3,485 MW |
| Equinix | 658 | 7,058 MW |
| Amazon Web Services | 504 | 14,070 MW |
| Microsoft | 301 | 11,090 MW |
| Equinix, Inc. | 222 | — |
| 206 | 7,109 MW | |
| Meta | 154 | 15,773 MW |
| Vantage Data Centers | 154 | 6,201 MW |
| NTT | 152 | 1,012 MW |
| DataBank | 149 | 1,055 MW |
Top 10 by deal value. Source: DC Hub deals tracker (1,972 historical deals · 144 in window).
| Date | Buyer | Seller | Value | MW |
|---|---|---|---|---|
| 2026-04-27 | Anthropic | $40.00B | — | |
| 2026-05-29 | Blackstone | $36.00B | — | |
| 2026-04-24 | Ares | ? | $13.60B | — |
| 2026-03-01 | Meta | ? | $10.00B | 1,000 MW |
| 2026-05-21 | Blackstone | $5.00B | — | |
| 2026-05-21 | Blackstone | $5.00B | — | |
| 2026-05-21 | Blackstone | $5.00B | — | |
| 2026-05-21 | Blackstone | $5.00B | — | |
| 2026-04-20 | Amazon | Anthropic | $5.00B | — |
| 2026-05-21 | Blackstone | $5.00B | — |
Auto-detected from news ingest. Live feed at /hyperscaler-deals.
| Detected | Actor | Value | Headline |
|---|---|---|---|
| 2026-05-25 | Anthropic | $25.0B | Amazon expands Anthropic partnership with $25 billion investment |
| 2026-05-25 | $5.0B | Google and Blackstone plan US$5 billion AI cloud venture | |
| 2026-05-25 | SUSE | $6.0B | SUSE's sovereignty pitch meets an inconvenient $6 billion question |
| 2026-05-25 | Microsoft | $25.0B | Microsoft lifts 2026 AI spend by $25 billion to cover component price rises |
Every number above is reproducible from public sources:
discovered_facilities table — aggregated from EIA-860, HIFLD,
ArcGIS FeatureServers, PeeringDB, OSM, operator filings, county permitting data.Every number on this page is also at GET /api/v1/reports/quarter.json.
Free, CC-BY-4.0. Cite us, embed us, or pipe us into your own quarterly research.
DC Hub. (2026). Quarterly Deep Report. https://dchub.cloud/reports/quarterly-deep. Licensed CC-BY-4.0.