Cloverleaf Infrastructure, a Houston-based developer of powered data-center sites, announced a strategic partnership with NVIDIA on August 21, 2026, with NVIDIA taking a minority investment in the two-year-old company to support the build-out of what both firms call AI factories. The partnership announcement did not disclose the size or terms of the investment.
The deal pairs a site developer with the company whose chips fill the buildings. Cloverleaf finds and prepares land with access to power, then delivers shovel-ready sites to the companies that construct and operate data centers. NVIDIA will work with Cloverleaf on the foundational infrastructure layer, and Cloverleaf will adopt the NVIDIA DSX platform to bring site, power, cooling, computing and facility decisions together earlier in project design.
“AI is accelerating demand for digital infrastructure at an unprecedented scale and driving strong economic development opportunities for American communities,” said David Berry, Cloverleaf’s co-founder and CEO. “This partnership strengthens Cloverleaf’s ability to identify and develop high-quality sites that provide the compute power for our country’s growth and security.”
NVIDIA’s Nico Caprez, vice president of global AI infrastructure growth, framed the deal around the physical inputs. “AI factories are the infrastructure of the intelligence age, and land, power and shell are their foundation,” he said in the release.
What Cloverleaf Actually Sells
Cloverleaf occupies a specific niche in the data-center supply chain: the powered-land layer. It does not operate data centers or sell computing capacity. It assembles sites, land plus a credible path to electricity, and hands them to the technology companies that build the facilities. According to its announcement, the company has delivered multiple gigawatt-scale projects to customers across North America since its founding in 2024.
Access to powered, shovel-ready sites is critical to bringing new capacity online quickly, per the announcement. Utilities’ interconnection queues, transmission upgrades, and local permitting set the pace of new data-center supply more than any hardware roadmap, which is why a developer that controls entitled, powered sites is strategically useful to a chip vendor trying to pull its next platform generation into deployed capacity faster. The pressure is visible in utility regulation too: the Tennessee Valley Authority’s board created a dedicated data-center rate class this month to keep AI-driven power costs off household bills.
Cloverleaf was formed in 2024 with an initial investment from Sandbrook Capital and NGP Energy Capital, according to the company’s own account. J.P. Morgan Securities served as exclusive financial advisor to Cloverleaf on the NVIDIA transaction, with Kirkland & Ellis as legal counsel.
Where DSX Fits in the Build Sequence
The technical substance of the partnership is Cloverleaf’s adoption of NVIDIA DSX, the AI-factory design and operations platform NVIDIA introduced at GTC Taipei on May 31, 2026. DSX bundles reference architectures, a simulation layer, and open-source operations software into a common framework NVIDIA describes as a playbook for infrastructure builders, covering everything from cluster design down to power, cooling and civil engineering.
For a site developer, the relevant pieces are the design-phase tools. Cloverleaf says it will use DSX to evaluate infrastructure tradeoffs against power, water and grid constraints before construction begins, and that DSX infrastructure and software will help customers optimize energy use and computing capacity once facilities are operating. The platform’s DSX MaxLPS software targets token output per megawatt within a fixed power budget (NVIDIA says its cooling and power-management techniques let operators run up to 40% more GPUs at their most energy-efficient operating point), while DSX Flex connects facilities to grid services so workloads can respond to utility signals like demand-response events. NVIDIA’s cloud partners, including CoreWeave, IREN and Nebius, are already deploying core DSX components, per the DSX platform announcement.
NVIDIA’s Expanding Infrastructure Balance Sheet
The Cloverleaf stake is one entry in a widening pattern of NVIDIA putting capital into the physical layer its GPUs depend on. On May 7, 2026, NVIDIA struck a comparable partnership with IREN, the data-center operator, targeting up to 5 gigawatts of DSX-aligned infrastructure and giving NVIDIA a five-year right to purchase up to 30 million IREN shares at $70 each — a potential $2.1 billion investment. Earlier this month NVIDIA also guaranteed up to $105 billion for an 8-gigawatt Ohio campus leased by OpenAI.
The logic is consistent across the deals: NVIDIA’s next platform generations sell into capacity that doesn’t exist yet, so the company is seeding the developers, operators and financing vehicles positioned to build it. A minority stake in the powered-land layer extends that strategy to the very start of the supply chain — the site itself, before a single rack is designed. Neither company disclosed a timeline for the first jointly developed site.
