Digital Realty's Q1 2026 Hyperscale Market Demand & Backlog Insights

Demand signals for digital infrastructure in hyperscale markets (DLR, Q1 2026)

1) Leasing momentum shows hyperscale demand converting into commitments

• Digital Realty reported “over 700 million” of new leases in Q1 2026 (and 423 million at DLR’s share), describing the quarter as the second-highest leasing quarter and nearly 70% above the prior next-highest quarter. ¹
• In hyperscale-relevant product lines, management highlighted strength across categories, including record zero-to-one megawatt plus interconnection bookings and stronger greater-than-a-megawatt leasing. ²
• Q1 also included a clear hyperscale milestone: a 200 megawatt AI inference-oriented lease with a AA-rated hyperscaler in Charlotte, described as the largest lease in company history and the first hyperscale deployment in that market. ³

2) AI-oriented requirements are a measurable portion of near-term demand

• Management stated that 21% of zero-to-one megawatt bookings were AI-oriented requirements (within the zero-to-one megawatt plus interconnection category). ¹
• They also characterized demand as shifting from “pilot to production”, particularly for inference, tying this to the 200 megawatt build and noting inference adoption is being driven by requirements moving beyond experimentation. ⁴

3) The “interconnection + proximity” product set is capturing demand relevant to hyperscalers

• Interconnection bookings totaled 186 million, 24% higher than a year ago, and were said to be led by APAC and North America, driven by demand for bulk fiber and ServiceFabric products. ⁵
• A record 98 million of leasing occurred within zero-to-one megawatt plus interconnection, described as a record within the past year’s cadence and supported by proximity, connectivity, and access to enterprises/service providers. ¹

4) Customer behavior signals scale-up (not just experimentation)

• Management noted that in the enterprise segment (adjacent to hyperscale demand), customers are migrating to larger committed capacity blocks, which they framed as important for delivering scaled inference. ⁴
• For hyperscale-sized deployments, they cited evidence of “large AI workloads” moving into scaled production environments, pointing to 10-plus megawatt leases in Dallas, São Paulo, and Tokyo during the quarter. ⁶

---

Backlog signals (visibility into future hyperscale buildout)

1) Backlog reached a new record and is sizable relative to annual lease commencements

• DLR reported total backlog at Q1 end reached 1.8 billion (new record), and 1.0 billion at DLR’s share. ⁵
• At DLR’s share level, 423 million of new bookings exceeded 204 million of commencements, implying backlog growth via net additions. ⁵

2) Backlog timing indicates sustained demand through multiple years

• Management guided to 44 million of leases scheduled to commence ratably throughout 2026, plus 247 million in 2027 and 242 million in 2028 and beyond (at DLR’s share). ⁵
• They also emphasized that their scaling backlog improves visibility over the long term, supporting strong, sustainable growth. ⁵

3) Backlog supports pipeline conversion capacity (a key “demand durability” signal)

• Management tied leasing strength to development conversion: they said development pipeline scaled to 16.5 billion at 100% share and that nearly 80% of this volume is in the Americas, aligned with “AI-oriented workloads from our largest customers.” ⁷
• They also stated that Q1 development ramp was supported by leasing and sales-funnel momentum, with 63 megawatts delivered (and 84% preleased) and 464 megawatts started (nearly 50% preleased), taking total capacity under construction to 1.2 gigawatts. ⁷

---

Bottom-line interpretation for hyperscale markets during Q1 2026

• Demand durability: Q1 showed a rare combination of (a) record interconnection and zero-to-one activity, (b) hyperscaler-scale megawatt leasing including the 200 MW AI inference lease, and (c) a stated shift from pilot to production for inference workloads. ⁵¹⁴³
• Backlog visibility: A new record backlog of 1.8B (1.0B at DLR share) plus a multi-year commencement schedule (2026, 2027, 2028+) indicates that Q1 leasing is not just “headline demand,” but is translating into measurable forward occupancy/earning visibility. ⁵
• Signal of hyperscale “scale-up”: Management explicitly linked large leasing to accelerating pace of large AI workloads into scaled production, and also highlighted that zero-to-one AI-oriented requirements are already a meaningful share of bookings. ¹⁶

(All figures and statements above are taken directly from the provided Q1 2026 Digital Realty earnings transcript excerpts.)