Ultrafast inter-site networking from WhiteFiber, DriveNets, and WEKA

Industry’s First Commercial Deployment of a Long-Distance, Scale-Across AI Supercluster

DriveNets AI Fabric connects two WhiteFiber H200 GPU clusters data centers 52 miles apart into a single GPU supercluster, freeing AI infrastructure from single-site power constraints while delivering 111.2 Tbps of bandwidth with sub-millisecond latency.

 

Full transcript

WhiteFiber just validated 111.2 terabits per second across 83 kilometers of dark fiber within a round trip of 0.9 milliseconds.

That’s within 8% of the physical limit for light in fiber. To demonstrate the performance of this innovation, we built one single GPU supercluster that acts as a single logical system across two separate geographical sites. Our networking partner, DriveNets Ethernet Fabric, and our storage partner, Weka’s Neuromesh Storage, were both integral to this achievement.

The result: enterprise AI workloads that exceed the scale, resilience, and compliance capabilities of any single facility.

In addition, we’ve built a solution that removes bandwidth constraints in AI infrastructure design and deployment. Patent applications have been submitted. Commercial launch is in Q3 2026.