
VAST Data, an AI data platform provider, has deepened its partnership with Google Cloud, adding to its menu of cloud-based tools for analytics and training to enhance AI development across multi-cloud and on-premise environments. This latest collaboration, which reflects the rising importance of AI tools, builds on an agreement between the two companies initiated in 2024.
In todayโs IT environments, AI is not only more central but far more complex, with a greater need to transform and transfer data from an array of previously siloed repositories to support AI tasks like large language model customization. VAST Dataโs architecture is built for this. Designed as the equivalent of an AI infrastructure operating system, its stack includes NMVe over RDMA connections to GPU and x86 processors, a modular storage menu, tools for Retrieval Augmented Generation (RAG) and an AI-native design. VAST competes with the likes of Snowflake for data warehousing capabilities, and is known for the top performance of its data storage tools.
VASTโs emphasis is on open systems that interoperate across all the major cloud providers. VAST version 5.2 is available on AWS, and on Azure, the platform is offered as a Directed Availability release, so Azure customers can use VASTโs disaggregated storage processing. It is on the Google platform that the VAST integration is the most complete.
Sameet Agarwal, VP and GM of Storage at Google Cloud, noted that โOur collaboration with VAST Data brings together Google Cloudโs robust infrastructure with a unified platform tailored for AI. This partnership simplifies operations and boosts performance for demanding AI workloads.โ
The list of VAST tools now available on Google Cloud includes the VAST InsightEngine for Intelligent Data Computation, which allows stateless compute and database services to feed directly into the data platform. This helps developers and IT pros more quickly run AI pipelines natively, which streamlines workflow and supports projects involving RAG, indexing and preprocessing.
VASTโs Exabyte-Scale Global Namespace lessens the borders between data siloes that typically occur when companies build hybrid and multi-cloud platforms that combine numerous providers. It offers a unified namespace for all the applications and services, and it connects on-premise, edge and cloud platforms for data access and mobility.
The companyโs Unified File, Object, Block and Database Access performs the heavy lifting required to bring object, file (NFS, SMB) and database tools into a unified, manageable system. Significantly, when all of these disparate data formats and services are combined, enterprise teams can more flexibly analyze and build AI models.
Most central to VASTโs integration with Google is its so-called disaggregated, share-everything architecture, which allows compute to scale from a base storage platform to enable fast throughput workloadsโessential for real time AI-based data usage. Supporting this is VASTโs Similarity Reduction tool, which lowers data handling costs.
In sum, the fact that a cloud leader like Google collaborates with VAST, which focuses on fast and open data transfer and management, demonstrates the pressing need for businesses to streamline data to build todayโs all-important AI deployments.
โThis collaboration represents a leap forward for organizations pushing the boundaries of whatโs possible with AI,โ said Renen Hallak, CEO of VAST Data. โBy combining Google Cloudโs elasticity with the VAST Platformโs intelligence, we empower developers to build smarter and scale faster.โ