Corvex has signed a multi-year agreement to deploy Nvidia's Blackwell GPU clusters, expanding its AI infrastructure at scale. The deal marks another major vote of confidence in Blackwell, Nvidia's latest flagship architecture for data center workloads.
The deployment includes Quantum-2 InfiniBand networking and dedicated high-speed storage infrastructure. Corvex completed its liquid-cooled GPU rollout in roughly two weeks, a tight timeline that signals both preparation and confidence in the setup. The expansion is funded through debt, customer prepayments, and existing cash rather than new share issuance, preserving equity for existing holders.
Nvidia benefits from yet another enterprise-scale Blackwell deployment as data center operators continue racing to lock in GPU capacity. The broader trend echoes across infrastructure plays, where power grid constraints are already forcing some regions to pump the brakes on expansion. For Corvex, the move is a bet that Blackwell workloads justify the capital outlay, with customer prepayments offsetting cash burn upfront.
