WhiteFiber, Inc. launched WhiteFiber Continuum to create distributed GPU superclusters across separate data centers. The company claims this is the first commercially available architecture of its kind. It addresses the rising demand for large-scale AI training and inference workloads.
The system connects two data centers located 83 kilometers apart. The link delivers 136 Tbps of bandwidth. The architecture maintains extremely low latency.
WhiteFiber developed the offering in collaboration with DriveNets and WEKA. The product allows regulated industries to pool GPU resources while maintaining data sovereignty. This launch marks WhiteFiber's expansion into the competitive AI infrastructure market.