India, Nov. 3 -- Vertiv Holdings and NVIDIA have introduced an AI factory approach that aims to reduce the time needed to move large-scale AI infrastructure projects from planning to productive capacity. Announced at NVIDIA GTC Berlin, the approach combines pre-engineered reference designs, virtual validation and factory-built deployment.

Under a defined model, the companies estimate that the approach could deliver up to 30% more cumulative token output in the first 12 months, using the same installed capacity. The improvement comes from bringing infrastructure online sooner rather than increasing the amount of installed compute.

For AI factory operators, securing GPUs and other compute resources is only one part of the challenge. Power...