NVIDIA DGX B200
8 × B200 per node, with 8 compute rails at 400G. That is 3.2 Tb/s of scale-out network per node.
- GPUs per node8 × B200
- Compute rails8 × 400G
- Scale-out bandwidth3.2 Tb/s
- Storage links2 × 400G
- In-band links2 × 100G
- BMC ports1
- Rack units10U
- Typical power14.3 kW
- Maximum power14.3 kW
What this means for the fabric
8 compute NICs means 8 rails. In a rail-optimized design each of those rails is sized on its own against the node count, then multiplied back up, so the leaf tier is driven by the rail count and not by the total endpoint count.
Storage and in-band management are separate fabrics with their own switches and their own optics, and the 1 BMC port per node feed the out-of-band network, which also has to carry every switch and every PDU in the cluster.
Power and racks
At 14.3 kW typical draw, this node fills a rack's power budget long before it fills the space. A 45 kW rack takes 3 of them on power alone, against 4 on rack units. Power is what actually sets the rack count.
Notes
Blackwell generation. Higher draw per node cuts rack density, so the same GPU count needs materially more racks than an H100 build.
Source: NVIDIA DGX B200 documentation, verified 2026-08. 10RU and ~14.3 kW confirmed. Confirmed 2026-08-06