DDN, global leader in AI and data intelligence solutions, announced at GTC Berlin 2026 that European AI cloud provider Nebul is using DDN Infinia to accelerate AI inference for sovereign workloads across its NVIDIA-powered infrastructure. Specifically DDN Infinia used to store and reuse key-value cache reducing GPU work required to serve AI models.
In testing DDN helped Nebul sovereign AI workloads find answers faster at lower cost nearly 10x faster time to first token, up to 5x faster token responses, 2.5x more input tokens per second, nearly 60% lower cost per million input tokens by reloading context instead of recomputing.
"Every time a GPU rebuilds context it has already processed, it is spending cycles repeating work," said Alex Bouzari, CEO and Co-Founder at DDN. "By making that context available across Nebul's GPU fleet, Infinia helps deliver faster inference and more productive use of GPU capacity while keeping data within the sovereign environment."
Keeping AI Context Available Across GPU Fleet DDN Infinia extends Nebul data controls to KV cache used during inference tenant subtenant dataset scope remain attached to stored context while cache capacity managed without interrupting inference. In multi-tenant environments DDN Infinia helps Nebul restore model context across GPU fleet stores KV blocks outside individual GPU servers so previously generated context restored when key remains available, keep tenant data isolated scope remain part of key namespace so saved context stays isolated, manage cache pressure without stopping inference applies eviction policies at dataset tenant level as capacity fills allowing inference continue while older cache removed.
"AI sovereignty should not come at the expense of performance," said Arnold Juffer, CEO of Nebul. "With DDN Infinia, we can reuse model context across our GPU infrastructure instead of spending valuable GPU cycles recomputing it. That makes inference faster and more efficient while keeping the complete AI workload within the sovereign environment."
Nebul customers illustrate types of regulated proprietary workloads driving demand for sovereign infrastructure Netherlands-based Centric supporting approximately 70% of Dutch municipalities uses Nebul to run sovereign AI inference with enterprise-grade security observability keeping workloads within European infrastructure. Utrecht-based Leukeleu runs private AI assistant SureGPT on Nebul while keeping each customer data under that customer control for internal documents board strategies financial models product roadmaps proprietary information.
