Cloudera has announced a strategic partnership with VAST Data to deliver a unified “AI Factory” architecture, aiming to solve a problem the companies say is holding back enterprise AI deployment across Asia Pacific: GPU starvation, where expensive accelerator clusters sit idle waiting for data.

The joint solution combines Cloudera’s next-generation lakehouse data services with the VAST AI Operating System, integrating the NVIDIA AI Data Platform reference design. It is designed to keep GPUs continuously fed with high-throughput, low-latency data across both on-premises environments and public clouds.

Addressing a costly infrastructure bottleneck

As organisations move to deploy generative and agentic AI, many are finding that older architectures were not built to support continuous AI pipelines spanning data preparation, training, inference and analytics. The result, the companies said, is GPU starvation — a scenario that can leave costly accelerator hardware underused while it waits on data.

Cloudera’s lakehouse architecture provides containerised data services — including data engineering, streaming, analytics and machine learning — deployable consistently across hybrid and multi-cloud environments. VAST’s Disaggregated Shared Everything architecture underpins its AI Operating System, delivering exabyte-scale data infrastructure and integrating vector database services with NVIDIA cuVS for GPU-accelerated indexing and search.

Together, the companies say the partnership delivers a unified AI Factory architecture from raw data ingestion through to model deployment, consistent operations across data centres and clouds, and improved GPU utilisation for structured, unstructured and multimodal datasets, with enterprise governance and compliance built in.

“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.” – Abhas Ricky, Chief Business Officer and General Manager, Applied AI at Cloudera.

An Asia Pacific-focused push

The companies frame the launch as particularly relevant to the region, where AI infrastructure investment is accelerating alongside constraints on compute and energy.

“AI infrastructure investment is surging across the Asia Pacific region, with numerous technology companies increasing their AI-related capital spending this year. Given the rising demand for compute power and heightening energy constraints, businesses must ensure their GPUs are fully utilized.” – Remus Lim, Senior Vice President, Asia Pacific at Cloudera.

VAST Data’s Vice President of APJC, Sunil Chavan, added that the regional AI opportunity is increasingly about activating the data enterprises already hold, arguing that combining the two companies’ platforms — which together account for many exabytes of customer-managed data — can help organisations turn existing data estates into a foundation for production-scale AI.

The partnership also extends to private and sovereign AI deployments, combining NVIDIA AI infrastructure and enterprise software with the VAST AI Operating System and Cloudera’s data and AI services. The Cloudera AI Inference Service, accelerated by NVIDIA NIM microservices, is one of the solutions enabling organisations to deploy and scale models including NVIDIA’s latest Nemotron open models.

The joint Cloudera–VAST AI Factory solution is available immediately through both companies’ enterprise sales teams and partner ecosystems, with reference architectures and industry-specific deployment patterns expected to expand through the rest of 2026.