Hut 8 Locks $9.8 B AI Data‑Center Lease – The Texas‑based energy‑infrastructure firm announced a 15‑year, $9.8 billion lease for an additional 352 MW of AI compute capacity at its Beacon Point campus, effectively doubling the tenant’s footprint to 704 MW and marking the first fully commercialized gigawatt‑scale AI data center in the United States.
Hut 8 Corp. (Nasdaq: HUT, TSX: HUT) has moved from a construction‑phase project to a revenue‑generating asset with the signing of a second long‑term lease for its Beacon Point campus in Nueces County, Texas. The agreement, valued at $9.8 billion over 15 years, adds 352 MW of IT capacity to an existing 352 MW lease signed in Phase 1, bringing the tenant’s total contracted power at the site to 704 MW. The campus already holds a 1,000 MW utility interconnection secured through AEP Texas, meaning the new lease fully utilizes the power envelope without requiring additional grid upgrades.
The tenant—described in the filing as a “high‑investment‑grade” counterparty—has chosen to double its footprint under the same triple‑net (NNN) lease terms that governed the inaugural agreement. The lease also embeds a 3 % annual base‑rent escalator and provides three optional five‑year renewal windows that could lift the campus‑level contract value to $50.2 billion if fully exercised. Expected net operating income (NOI) from the Phase 2 lease alone is projected at $655 million per year once stabilized, pushing the average annual NOI for the entire 1,000 MW campus to roughly $1.31 billion.
Power‑First Development Model
Hut 8’s “power‑first” approach—securing grid capacity before building the physical facility—has proven decisive in a market where power constraints often throttle AI expansion. By locking in 1,330 MW of utility capacity (including 500 MW earmarked for the new AI factory) ahead of construction, Hut 8 mitigated a common risk that has slowed competitors such as Equinix and Digital Realty in their AI‑focused projects. The design leverages NVIDIA’s DSX reference architecture, a blueprint that enables higher density AI workloads while preserving energy efficiency.
The redesign of the original Phase 1 data hall to the DSX standard boosted IT capacity by 57 % within the same footprint, a move that aligns with Gartner’s 2023 forecast that AI‑optimized data centers will achieve 30‑40 % higher utilization than traditional facilities. By replicating that architecture in Phase 2, Hut 8 not only satisfies the tenant’s immediate demand but also positions the campus to accommodate future AI models that require more GPU‑dense configurations.
Financial Discipline and Capital Allocation
The lease adds $9.8 billion of base‑term contract value to Hut 8’s balance sheet, raising the cumulative contractual value across its portfolio to $26.6 billion. All of the company’s AI data‑center contracts are backed by investment‑grade counterparties, a factor that Bloomberg’s credit analysts cite as a key differentiator for infrastructure REITs. The transaction also dovetails with a $250 million stock repurchase program launched in December 2024, signalling confidence from the board that the firm can return capital to shareholders while still funding its pipeline.
Implications for the AI Infrastructure Market
The announcement arrives at a pivotal moment for AI infrastructure. IDC predicts that global AI‑focused data‑center spending will exceed $200 billion by 2027, driven largely by the surge in large language model (LLM) training and inference workloads. Hut 8’s ability to deliver a gigawatt‑scale, power‑secured campus in under two years challenges the traditional “build‑then‑sell” timeline that has hampered many providers.
Compared with rivals, Hut 8’s model reduces the capital intensity of each megawatt. While Equinix’s AI‑specific sites typically require $1.5–$2 million per MW of power‑ready capacity, Hut 8’s power‑first strategy reportedly brings that figure closer to $1 million per MW, according to an internal cost model cited by the company. This cost advantage could translate into lower lease rates for tenants, potentially accelerating AI adoption among mid‑size enterprises that have been price‑sensitive.
Enterprise Marketing Teams Take Note
For B2B marketers, the fully commercialized Beacon Point campus offers a concrete example of how AI infrastructure can be packaged as a service offering. The triple‑net lease structure shifts operational risk to the tenant, allowing enterprises to focus on model development rather than data‑center management. Moreover, the guaranteed power supply and NVIDIA‑certified architecture provide a compelling value proposition for firms looking to run LLM inference at scale without the overhead of building their own facilities.
The marketing teams can leverage the predictable cost structure to embed AI‑driven services into their go‑to‑market strategies, while the underlying power reliability ensures consistent performance for high‑throughput workloads.
Market Landscape
- Power Availability – AEP Texas’s 1,000 MW interconnection gives Hut 8 a competitive edge over sites that still rely on incremental grid upgrades.
- Hardware Standardization – By aligning with NVIDIA DSX, the campus reduces integration friction for AI developers accustomed to the same stack used in Google’s TPU‑backed pods.
- Financial Predictability – Triple‑net leases with built‑in escalators provide a revenue stream that is less volatile than spot‑market power contracts, a factor highlighted in a recent Forrester report on AI‑infrastructure investment risk.
Top Insights
- Hut 8’s power‑first model cuts AI‑data‑center build time to under two years, outpacing most competitors.
- The $9.8 billion lease pushes the campus’s average annual NOI to $1.31 billion, establishing a new benchmark for AI‑focused REITs.
- NVIDIA DSX‑based design delivers 57 % more compute density, enabling faster LLM training cycles for enterprise customers.
- Triple‑net lease terms with 3 % annual escalators create a predictable cost structure that can be built into long‑term AI‑budget forecasts.
- The campus’s 1,000 MW power cap, already secured, eliminates a common bottleneck that has delayed AI deployments at rival sites.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
