Partnership will integrate VAST AI OS into Hyperfusion’s modular Sentinel data centre, enabling sovereign AI inference closer to available power, enterprise data and end users.
VAST Data and Hyperfusion are joining forces to accelerate AI infrastructure deployment, combining a unified data platform with modular data centre infrastructure designed to bring production AI inference online in weeks rather than years.
Announced on October 20, 2026, the planned integration will make VAST AI Operating System (AI OS) the preferred data platform for Hyperfusion’s Sentinel modular AI infrastructure. The companies aim to help governments and enterprises deploy sovereign, confidential AI environments while retaining control over their data and operating requirements.
The partnership addresses a growing constraint in AI expansion: securing GPUs is not always the biggest challenge. Power availability, cooling capacity and suitable data centre space can delay deployments, forcing organisations to rethink where AI infrastructure can be installed and operated.
Abu Dhabi-headquartered Hyperfusion addresses this challenge through factory-built modular infrastructure. Its Sentinel platform packages a complete AI data centre into a transportable 45-foot ISO High Cube container, with a reference configuration of 96 GPUs and approximately two to three petabytes of data capacity. HyperPod extends the model to larger, multi-megawatt deployments, with each pod scaling to approximately 20 petabytes.
“Power isn’t arriving in neat, hundred-megawatt blocks on the timetable AI demands.” — Jeff Denworth, Co-Founder, VAST Data
The combined offering is intended for government, enterprise, telecommunications and energy customers that need to deploy AI rapidly or keep workloads and data within defined national or organisational boundaries.
Sentinel will support confidential computing to protect models and data while in use, and can operate in fully air-gapped environments where required. Customers can purchase the infrastructure outright or consume it as a managed service. Integrating VAST AI OS is expected to provide a common data architecture across deployments, with a global namespace spanning Sentinel and HyperPod environments.
“Every government and enterprise we speak to wants the same three things: capacity now, data that never leaves their control, and a platform that grows without starting over,” said Quentin Reyes, Founder and CEO of Hyperfusion. He said factory commissioning and confidential computing are designed to help customers bring AI infrastructure online quickly while maintaining control of sensitive models and data.
The partnership also reflects the growing need to position inference infrastructure closer to users, enterprise data and available power. As multimodal AI, computer vision and physical AI expand, organisations face increasing demands for both compute and data capacity.
“The AI infrastructure conversation has been dominated by how many GPUs you can secure, when increasingly the gating factor is where you can actually put them,” said Jeff Denworth, Co-Founder of VAST Data. He said the collaboration aims to make the data layer as portable as the compute infrastructure, avoiding isolated data environments every time a new deployment is established.
VAST Data and Hyperfusion will begin joint go-to-market activities and customer introductions while completing technical validation of VAST AI OS within Sentinel and HyperPod architectures. The companies also plan to explore a common control plane to manage distributed Sentinel deployments across multiple sites.
The initiative positions modular infrastructure and portable data services as potential alternatives to conventional data centre expansion, particularly for organisations seeking faster AI deployment without compromising data sovereignty or operational control.
