🔥Stalecollected in 5m

Nebius to Raise $3.75B Convertible Notes

PostLinkedIn
🔥Read original on 36氪

💡$3.75B for AI data centers & GPUs: key infra boost for scaling models

⚡ 30-Second TL;DR

What Changed

Private issuance of $2B 2031 notes and $1.75B 2033 notes

Why It Matters

This funding bolsters Nebius' ability to scale AI compute resources, potentially lowering costs for AI workloads and enhancing competition in cloud AI services.

What To Do Next

Check Nebius AI cloud pricing for GPU-heavy training workloads.

Who should care:Enterprise & Security Teams

🧠 Deep Insight

Web-grounded analysis with 6 cited sources.

🔑 Enhanced Key Takeaways

  • Nebius originated from Yandex N.V., which divested its Russian assets for $5.4 billion in mid-2024 and rebranded as Nebius, resuming Nasdaq trading in October 2024 with assets including a Finnish data center and 1,300 relocated engineers.[1][2]
  • Nebius offers GPU-as-a-Service with NVIDIA H100, H200, and Blackwell B200 clusters, proprietary server racks, InfiniBand networking, and 'Nebius Fabric' for low-latency multi-node AI training.[1][5]
  • The Mäntsälä data center in Finland features a heat-recovery system for local heating, achieving one of the industry's lowest PUE scores, and houses ISEG, the #19 most powerful supercomputer globally.[1][5]

🛠️ Technical Deep Dive

  • Deploys NVIDIA H100, H200, and Blackwell B200 GPUs in clusters supporting training and inference, integrated with pre-configured drivers, high-performance InfiniBand networking, and orchestration via Kubernetes or Slurm.[1][5]
  • Proprietary 'Nebius Fabric' enables seamless multi-node training for trillion-parameter models, with custom-designed server racks reducing latency.[1]
  • Mäntsälä facility uses heat-recovery for efficiency (low PUE), powering large-scale GPU superclusters built in-house.[1][5]

🔮 Future ImplicationsAI analysis grounded in cited sources

Nebius will expand to 75MW capacity in Finland by tripling current data center size
Plans announced to increase Finnish facility capacity and add co-locations in Paris and Kansas for lower latency and higher scale.[2]
Partnerships with Microsoft and Meta will drive multi-billion GPU contracts
Nebius has secured large contracts with these tech giants amid surging AI compute demand, plus Nvidia collaboration.[4]
Acquisition of Tavily enhances agentic AI capabilities
Tavily provides web search tools for autonomous AI agents, integrated into Nebius' full-stack platform for production-scale deployment.[4]

Timeline

2011-05
Yandex N.V. lists on Nasdaq as Dutch holding for Russian tech giant.
2022-02
Geopolitical events trigger Yandex N.V. asset divestment process from Russia.
2024-07
Yandex N.V. completes $5.4B sale of Russian business and rebrands to Nebius AI.
2024-10
Nebius Group resumes trading on Nasdaq as AI infrastructure provider.
2025-05
Selected by Israel Innovation Authority for national AI supercomputer infrastructure.

📰 Event Coverage

📰

Weekly AI Recap

Read this week's curated digest of top AI events →

👉Related Updates

AI-curated news aggregator. All content rights belong to original publishers.
Original source: 36氪