GPU clusters sitting idle while waiting for data to arrive have become one of the more expensive problems in enterprise AI deployment. Cloudera, a data management and analytics platform company, and VAST Data, which builds storage and data infrastructure for large-scale AI workloads, have announced a partnership to address that bottleneck by connecting their platforms into a shared architecture designed to keep accelerator clusters continuously fed.
The combined system integrates Cloudera's containerized lakehouse services, covering data engineering, streaming, analytics, machine learning and AI, with the VAST AI Operating System, which unifies high-performance storage, a database layer and a global namespace for managing data across distributed environments. The architecture follows the NVIDIA AI Data Platform reference design and incorporates NVIDIA cuVS for GPU-accelerated vector indexing and search. Customers can run the solution on-premises, in private cloud or across public cloud environments from the same operational model.
Abhas Ricky, Chief Business Officer and GM of Applied AI at Cloudera, described the commercial problem the partnership is built around: "Enterprises are investing billions in GPUs, yet many struggle to achieve full utilization due to data bottlenecks. Our partnership with VAST eliminates GPU starvation and enables customers to build true AI factories—where data flows seamlessly from ingestion to insight."
The partnership brings together a combined 60 exabytes of customer-managed data across both companies' installed bases. VAST's storage layer uses what the company calls a Disaggregated Shared Everything architecture, which separates compute from storage while maintaining a unified namespace, enabling high-throughput, low-latency data delivery to GPU clusters running training and inference workloads.
Jeff Denworth, Co-Founder at VAST Data, framed the gap the joint platform closes for enterprise customers: "Most enterprises already have the data they need for AI. The challenge is unlocking the value in data to create a continuous pipeline of AI inference, fine-tuning, and data analysis to build the next generation of intelligent applications. Together, Cloudera and VAST are helping customers build AI factories that connect data, intelligence, and infrastructure into a single operational platform for AI across hybrid environments."
The joint solution also supports private and sovereign AI deployments, combining NVIDIA AI infrastructure, NVIDIA AI Enterprise software and the VAST AI OS for organizations in regulated industries that cannot route data through public cloud environments. Cloudera's AI Inference Service, accelerated by NVIDIA NIM microservices, lets those organizations deploy and scale models, including NVIDIA's Nemotron open models, within their own infrastructure. Apache Spark workloads can run with NVIDIA cuDF acceleration through VAST's high-throughput data services, extending GPU-accelerated processing into Cloudera's data engineering layer.
The joint solution is available immediately through both companies' enterprise sales teams and partner networks. Reference architectures, validated deployment patterns and industry-specific configurations are scheduled to expand through the remainder of 2026.
