DLR · Q2 2025 Earnings

The dominant tonal shift across the prepared-remarks narrative is that 0-1MW + interconnection has been promoted from "emerging opportunity" to lead growth engine. "Record performance in the zero to one megawatt plus interconnection product set stole the show in the quarter with $90 million of bookings…18 percent higher than our prior record set only two quarters ago. Over the past four quarters, we have booked over $300 million in this category, up from approximately $200 million in 2023." This matters because the retail/colo product carries higher returns and stickier economics than wholesale hyperscale — the bookings trajectory implies the revenue mix shift is real, not a one-quarter spike.

The second shift is on the funding model. Where prior commentary framed hyperscale growth as balance-sheet-constrained, management now describes a structural resolution: "The early success of our U.S. hyperscale fund improves our strategic position by enabling us to continue to meet the growing and diverse needs of our hyperscale customers without overtaxing our balance sheet." The U.S. Hyperscale Fund is oversubscribed and underwrites >$15B of additional development capacity. That removes the most common bear objection — that DLR has to choose between leverage and growth.

Third, management is explicitly trying to flatten expectations. "While the rate of acceleration and bottom-line growth this quarter is notable and demonstrates the significant momentum we have in our business, our growth will be best measured in years." Paired with a separate framing of the 5-7% growth range as a "consistent compounder" floor rather than ceiling, this reads as deliberate expectation management — guarding against quarterly beat-chasing while signaling confidence in the durability of the algorithm.

Fourth, on enterprise AI: management is candid that current contribution is modest and the workload is sitting in hyperscaler clouds today, but draws the cloud-to-hybrid analogy. The tone is patient rather than promotional — TAM expansion is treated as inevitable but not imminent.

Finally, hyperscale capacity-block discussions have moved out: "most of the discussions that we are having are focused on late 2026 and early 2027 deliveries." Power alignment is now framed as a competitive moat rather than a near-term bottleneck.