WhiteFiber Launches Cross-Data-Center GPU Supercluster to Break AI Infrastructure Barriers
Event summary
- WhiteFiber launched WhiteFiber Continuum, a distributed GPU supercluster architecture linking two data centers 83km apart with 136 Tbps bandwidth and 0.9ms latency.
- The solution leverages DriveNets AI Fabric and WEKA NeuralMesh for high-performance networking and data storage across sites.
- Continuum enables enterprises to pool GPU resources across locations, ensuring data sovereignty and resilience without single points of failure.
- The architecture is commercially available as of September 23, 2026, with patent applications submitted for the underlying implementation.
The big picture
WhiteFiber's Continuum addresses a long-standing constraint in AI infrastructure: geographic distance limiting cluster performance. By enabling distributed GPU superclusters, the solution could reshape how enterprises deploy AI workloads, particularly in regions with power or space limitations. The launch comes as demand for scalable, resilient AI infrastructure grows, with implications for cloud providers and hyperscalers facing similar challenges.
What we're watching
- Strategic Partnerships
- How WhiteFiber's collaboration with DriveNets and WEKA will evolve and whether it can sustain competitive differentiation.
- Market Adoption
- The pace at which enterprises will adopt Continuum, particularly in regulated industries requiring data sovereignty.
- Technical Scalability
- Whether the architecture can scale beyond the current 83km distance and handle even larger distributed workloads.
Related topics