Businesswire·2025-12-17 14:27Core Insights - Nebius has launched Nebius AI Cloud 3.1, enhancing its AI cloud platform with next-generation NVIDIA Blackwell Ultra compute and improved operational capabilities [1][2] - The new version focuses on providing operational visibility, resource planning, and addressing the needs of customers transitioning to large-scale AI adoption [2][4] Infrastructure and Performance - Nebius is deploying NVIDIA Blackwell Ultra infrastructure globally, being the first cloud provider in Europe to operate both NVIDIA GB300 NVL72 and HGX B300 systems in production [3] - The platform achieves leading results in MLPerf® Training v5.1 benchmarks, supported by hardware-accelerated networking and enhanced storage caching [3][9] Operational Transparency - Version 3.1 introduces Capacity Blocks and a real-time Capacity Dashboard, offering customers visibility into reserved GPU capacity and availability across data center regions [4][9] - Project-level quotas and new lifecycle object storage rules provide granular control over resource allocation and cost management [4][9] Ecosystem and Developer Usability - The Nebius AI Cloud ecosystem is expanding with new native integration with Dstack and simplified deployment of NVIDIA BioNeMo NIM microservices [5] - Enhanced developer usability features include improved Slurm-based orchestration and FOCUS-compliant billing exports [5][9] Security and Compliance - The latest release builds on Aether's security foundation with HIPAA-compliant audit logs, per-object access controls, and enhanced IAM with Microsoft Entra ID integration [6][12] - These features advance compliance capabilities for customers in regulated sectors and government [6][12]