VAST Data Mistral AI and Mistral Compute Strategic Alliance to Build NVIDIA Accelerated AI Factories and Large Scale Supercomputing Clusters in Europe
VAST Data has shared the particulars of its strategic alliance with Mistral AI and Mistral Compute, an alliance to build a new tier of NVIDIA accelerated AI factories across Europe. Mistral Compute selected VAST AI Operating System as the core data layer for its large scale NVIDIA GB300 NVL72 supercomputing cluster. A combination of three critical ingredients for enterprise grade intelligence, NVIDIA accelerated hardware, Mistral cutting edge language models and VAST converged data architecture has been unveiled.
The collaboration results in one of the highest densities of NVIDIA GB300 NVL72 computing capacity available in Europe. Mistral Compute manages the physical infrastructure and provides access, and VAST supplies the undergirding software engine to aggregate data access, transfer and governance across the entire distributed infrastructure. Purpose built to abolish storage silos and data pipelines, the platform is designed to offer researchers and automation engines a single namespace across all training, inference, and retrieval workflows.
For large scale supercomputing to be efficient and performant, it requires a specialized data layer that does not bottleneck the expensive GPUs. In order to support maximized throughput, this data layer needs to be capable of ingesting, storing, processing and serving a multitude of data types including, but not limited to: files, tables, vectors, event streams, persistent memory. VAST is already active in the cloud service providers that Mistral is using for their core model research, so data teams can migrate complex workloads to these distributed platforms without rewriting their data pipeline architecture.
A modern AI factory is a distributed system of compute, data, memory, and models, and the architecture has to be designed as 1 from the ground up. That means scaling linearly with the model, working across the sites and geographies a frontier program now spans, and handling every shape of data the system depends on, from KV cache to vectors to memory to the traces that feed the next round of training. With Mistral and NVIDIA we are moving from large GPU clusters to durable production platforms and the VAST AI OS is the data layer underneath.
The integration of frontier models into large enterprise and public sector workflows requires data sovereignty. This partnership supports that requirement by enabling enterprises to run Mistral models physically adjacent to where their local data storage resides, ensuring data privacy, governance, and centralized control over information at a unified interface that works across research, cloud services and on premises infrastructure.
Yann Lger, VP Cloud Compute, Mistral AI stated that purpose built solutions are a necessity for enterprises:
AI Infrastructure needs to be purpose built to serve the unique needs of large scale training and inference. With Nvidia GB300s and the VAST Data AI platform, Mistral is building frontier infrastructure to serve researchers, enterprises, and customers globally.
NVIDIA VP of Supercomputing EMEA region, Rod Evans, highlights the effort to provide local compute capabilities, with the joint offering aggregating accelerated hardware, leading edge models and unified data management to create a production ready platform enabling enterprises to deploy a scalable solution tailored to their long term operations and meet data sovereignty and residency requirements.
