Vast Data eyes new funding with Nvidia partnership

Vast Data, an Israeli data infrastructure firm, is deepening its partnership with Nvidia as it prepares to raise new capital at a valuation between US$25 billion and US$30 billion.

The company announced a fully CUDA-accelerated AI operating system at the VAST Forward 2026 conference, enabling its software to run directly on Nvidia-powered servers.

This update aims to streamline AI workflows by integrating storage, databases, analytics, and inference into a single platform, reducing data bottlenecks in large-scale deployments.

The new system is built around Vast’s CNode-X servers, which are Nvidia-certified and designed to run Vast’s software on GPU infrastructure.

Nvidia CEO Jensen Huang highlighted the system’s ability to handle longer, more complex AI tasks.

Vast’s close relationship with Nvidia includes partnership and investment, with the funding round expected to include Nvidia.

Founded in 2016, Vast employs around 300 staff in Israel and had about US$200 million in annual recurring revenue (ARR) as of early 2025, also achieving positive operating cash flow.

🔗 Source: Calcalist

🧠 Food for thought

Implications, context, and why it matters.

Behind the valuation and big customer deals plus fast growth

VAST secured a $1.17 billion agreement with AI cloud provider CoreWeave. CoreWeave will use VAST as its primary data platform 1.
Other AI players have also adopted the platform, including Elon Musk’s xAI and specialized cloud providers like Lambda Labs 2.
These deals pushed contracted annual recurring revenue above $500 million this year. That figure rose from the $200 million reported for early 2025, while total software bookings now exceed $4 billion 3.

The AI data stack is merging into one integrated platform

The collaboration marks a shift in AI infrastructure. Teams are moving away from piecing together separate storage, database, analytics, and AI inference stacks, toward a unified model 4.
VAST runs its software directly on GPU-accelerated CNode-X servers (Nvidia-certified servers designed to run Vast’s software on Nvidia-powered GPU infrastructure). It also embeds Nvidia software libraries for tasks like vector search, which narrows the gap between the data platform and the compute engine 5.
This design opens the door to a new class of closely linked AI systems. It also forces traditional enterprise storage vendors to pursue deeper co-engineering with GPU makers like Nvidia or risk becoming more commoditized in AI deployments 6.