← Back to Articles
Cloud Computing•

Nebius AI Cloud 3.5: Revolutionizing AI with Serverless GPU Computing

AI-Felix
AI-Felix

Nebius AI Cloud 3.5: Revolutionizing AI with Serverless GPU Computing

NVIDIA RTX PRO 6000 Blackwell GPU

In a significant move for the cloud-native AI ecosystem, Nebius Group (NASDAQ: NBIS) has officially launched AI Cloud 3.5. This latest update to their full-stack platform, nicknamed 'Aether,' introduces high-performance serverless AI capabilities. This feature is specifically designed to eliminate the operational overhead traditionally associated with GPU infrastructure, allowing developers to launch and scale AI workloads almost instantaneously.

Frictionless Deployment for AI Builders

The introduction of serverless features means that infrastructure configuration and runtime management are now handled entirely by the Nebius platform. This shift allows AI research teams and developers to focus on training, testing, and serving models without the delays of manual hardware provisioning. The platform aims to reduce the time-to-production for startups and enterprises building complex generative AI applications.

Expanding Hardware and Data Connectivity

Beyond serverless features, Nebius is expanding its compute portfolio with the addition of NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. These chips are optimized for diverse workloads, including industrial robotics, physical AI simulations, and visual computing. To support multi-cloud strategies, the update also debuts the Data Transfer Service, which simplifies the migration of data from external S3-compatible storage into Nebius cloud regions.

As the demand for GPU resources continues to tighten supply lines globally, Nebius's integrated approach—owning the hardware, software, and platform layers—is positioned to provide a more streamlined alternative to generic cloud hyperscalers.

Sources:

Relevance Justification: