NVIDIA unveiled NVIDIA Spectrum-XGS Ethernet, a cutting-edge technology that allows distributed data centers to operate as unified, giga-scale AI super-factories.
As AI adoption accelerates, individual data centers are nearing the limits of power and capacity within a single facility. To overcome these constraints, organizations must scale beyond the walls of a single building. Traditional Ethernet infrastructure, however, suffers from high latency, jitter, and unpredictable performance, making it unsuitable for such expansive growth.
AI Authority Trend: DataPelago Nucleus Outperforms Nvidia cuDF in New Benchmark Results
Spectrum-XGS Ethernet addresses these challenges by introducing scale-across infrastructure to the NVIDIA Spectrum-X Ethernet platform. This new capability acts as a third pillar of AI computing alongside scale-up and scale-out enabling multiple, distributed data centers to interconnect seamlessly. The result is the development of massive AI super-factories designed to deliver intelligence at an enormous, giga-scale level.
“The AI industrial revolution is here, and giant-scale AI factories are the essential infrastructure,” said Jensen Huang, founder and CEO of NVIDIA. “With NVIDIA Spectrum-XGS Ethernet, we add scale-across to scale-up and scale-out capabilities to link data centers across cities, nations and continents into vast, giga-scale AI super-factories.”
Fully built into the Spectrum-X platform, Spectrum-XGS Ethernet uses smart algorithms that automatically adapt to the distance between data center locations, ensuring smooth and efficient performance. It incorporates advanced distance-aware congestion control, precision latency management, and end-to-end telemetry, nearly doubling the performance of the NVIDIA Collective Communications Library. This acceleration of multi-GPU and multi-node communication ensures predictable performance across geographically distributed AI clusters, effectively allowing multiple data centers to operate as a single, high-performance AI super-factory optimized for long-distance connectivity.
AI Authority Trend: Virtana Partnering with NVIDIA to Enhance AI Factory Observability
Hyperscale innovators, such as CoreWeave, are among the first to adopt Spectrum-XGS Ethernet. “CoreWeave’s mission is to deliver the most powerful AI infrastructure to innovators everywhere,” said Peter Salanki, cofounder and CTO of CoreWeave. “With NVIDIA Spectrum-XGS, we can connect our data centers into a single, unified supercomputer, giving our customers access to giga-scale AI that will accelerate breakthroughs across every industry.”
The Spectrum-X Ethernet platform delivers 1.6x greater bandwidth density than conventional Ethernet, supporting multi-tenant, hyperscale AI operations, including the world’s largest AI supercomputer. Made up of NVIDIA Spectrum-X switches and NVIDIA ConnectX-8 SuperNICs, it delivers seamless scalability, ultra-low latency, and top-tier performance ideal for enterprises building the next generation of AI.
This announcement builds on NVIDIA’s recent breakthroughs in networking, such as Spectrum-X and Quantum-X silicon photonics switches. These technologies make it possible to link millions of GPUs across multiple locations, all while keeping energy use and operational costs low key for powering the next wave of AI factories.
AI Authority Trend: SK Telecom and VAST Data Partner for Korea’s Sovereign AI Cloud with NVIDIA Blackwell GPUs
To share your insights, please write to us at sudipto@intentamplify.com
