VAST Data today announced an expanded partnership with Google Cloud, enabling customers to deploy the VAST AI Operating System (AI OS) as a fully managed service and extend a unified global namespace across hybrid environments. Powered by the VAST DataSpace, enterprises can seamlessly connect clusters running in Google Cloud and on-premises locations, eliminating complex migrations and making data instantly available wherever AI runs.

Enterprises want to run AI where it performs best, but data rarely lives in one place and migrating can take months and costs millions. Fragmented storage and siloed data pipelines make it hard to feed the AI accelerators with consistent, high-throughput access and every environment change multiplies governance and compliance burdens.

VAST and Google Cloud address this challenge by making data placement a choice rather than a constraint. In this recorded demonstration, VAST showcased the power of the VAST DataSpace to connect clusters across more than 10,000 kilometres, linking one in the United States with another in Japan. This configuration delivered seamless, near real-time access to the same data in both locations while running inference workloads with vLLM, enabling intelligent workload placement so organisations can run AI models on TPUs in the US and GPUs in Japan without duplicating data or managing separate environments.

“Through our partnership with Google Cloud, we’re meeting customers where they are by delivering a fully managed AI OS,” said Jeff Denworth, Co-Founder at VAST Data. “When leveraging our global namespace with intelligent streaming, Google Cloud customers can auto-deploy a VAST-managed cluster via Google Cloud Marketplace and start production in minutes, providing integrated governance and billing, elastic scale, and supporting the hybrid cloud mission – and it’s all handled by VAST, making enterprise data instantly usable for agentic workloads.”

“Bringing VAST AI Operating System to Google Cloud Marketplace will help customers quickly deploy, manage, and grow the data solution on Google Cloud’s trusted, global infrastructure,” said Nirav Mehta, Vice President, Compute Platform at Google Cloud. “VAST can now securely scale and support customers on their digital transformation journeys.”

Powering Google Cloud TPUs with seamless data access and near-local performance

Recent performance results also show how the VAST AI Operating System connects seamlessly to Google Cloud Tensor Processing Unit (TPU) virtual machines, integrating directly with Google Cloud’s platform for large-scale AI. In testing with Meta’s Llama-3.1-8B-Instruct model, the VAST AI Operating System delivered model load speeds on par with local NVMe disks, while maintaining predictable performance during cold starts. These results confirm that the VAST AI OS is not just a data platform, but a performance engine designed to keep accelerators fully utilised and AI pipelines continuously in motion.

“The VAST AI OS is redefining what it means to move fast in AI. Loading models at speeds comparable to local NVMe while delivering a full suite of advanced platform features,” said Subramanian Kartik, Chief Scientist at VAST Data. “This is the kind of acceleration that turns idle accelerators into active intelligence, driving higher efficiency and faster time to insight for every AI workload.”

With VAST on Google Cloud, customers can benefit from:

• Deploy AI in Minutes, Not Months: Organisations can run production AI workloads on Google Cloud today against existing on-premises datasets without migration planning, transfer delays, or extended compliance cycles. Using VAST DataSpace and intelligent streaming, they can present a consistent global namespace of data across on-prem and Google Cloud instantly.
• Reduce Data-Movement Costs: Stream only the subsets that models require to avoid full replication and reduce egress – cutting footprint and redirecting budget from data movement to AI innovation with infrastructure that is future-ready for the demanding AI pipelines in genomics, structural biology, and financial services.
• Maximise Google Cloud Innovation with Flexible Data Placement: Choose what to migrate, replicate, or cache to Google Cloud while keeping one namespace and consistent governance by applying unified access controls, audit, and retention policies everywhere to simplify compliance and reduce operational risk. Leverage VAST DataStore and VAST DataBase to unify prep, training, inference, and analytics without rewiring pipelines.
• TPU-Ready Data Path: Feed TPU VMs over validated NFS paths with optimised model-load and small-file/metadata-aware I/O, achieving warm-start load times on par with local NVMe while maintaining predictable, steady behaviour during cold-starts.
• Build on a Unified Platform: The VAST AI Operating System delivers a DataStore, DataBase, InsightEngine, AgentEngine and DataSpace that scales across on-premises and Google Cloud environments and adapts to changing business needs without architectural rewrites, enabling data scientists to use a variety of access protocols with a single solution.

VAST can be deployed today in Google Cloud. Joint validation and reference guidance for establishing a VAST DataSpace spanning Google Cloud and external clusters are available to qualified customers and partners.