Corvex signs multi-year deal to supply NVIDIA Blackwell GPU clusters

Corvex, an engineering-led AI computing platform company, said it has signed a multi-year agreement to provide clusters of NVIDIA Blackwell GPUs. The announcement, carried by Yahoo Finance, positions Corvex as a provider of GPU-accelerated infrastructure for organizations running AI workloads that require significant compute capacity.

Blackwell is NVIDIA’s data-center GPU platform, designed to serve training and inference workloads for large-scale AI systems. By tying its multi-year supply arrangement to NVIDIA’s latest GPU generation, Corvex is indicating a commitment to building out delivery capacity for customers that want to deploy NVIDIA’s newer AI compute roadmap rather than relying on older architectures.

The deal is framed as an agreement to “provide clusters” rather than a one-off component sale. In practice, a GPU cluster typically means an integrated set of compute and networking resources configured to run AI software stacks efficiently, reducing the work customers must do to assemble and validate high-performance AI infrastructure.

Corvex’s business model centers on engineering and deployment of GPU-accelerated infrastructure. The company, which trades on the Nasdaq under MOVE, has described itself in the AI computing space as delivering infrastructure to support AI workloads. The agreement with NVIDIA, if executed as stated, would support Corvex’s ability to offer packaged GPU clusters to customers seeking near-term capacity.

NVIDIA, for its part, benefits from customer-facing distribution channels that can translate its chip roadmap into ready-to-run data-center systems. Multi-year arrangements also tend to provide companies like NVIDIA with clearer visibility into demand from infrastructure builders, while giving those builders a more stable basis to plan procurement and delivery.

Still, the announcement does not spell out the commercial details investors usually look for. It does not disclose financial terms, the number of systems or GPUs covered, delivery timing, or whether the clusters are intended for a single end customer or multiple deployments. It also does not provide any guidance on expected revenue impact for Corvex or incremental procurement expectations for NVIDIA.