The pod lands on land you already own and runs on a connection that already exists. It can be set up as a non-permanent structure, and when the main build is ready it leaves on the same truck it arrived on.
Each pod has one customer and runs as single-tenant bare metal, so you get the whole machine. Operating access down to BIOS, BMC, firmware and the fabric manager can be granted in the contract. Day to day the pod is run remotely, with contracted field service in your region.
Pacific's GPUs run as a managed cluster, or your own GPUs sit in a Pacific pod with power and cooling already built in. Either way you are the only tenant.
Pacific rents GPUs, builds managed clusters and sells pods to selected partners.
We work with AI labs, governments and defense that need compute on their own ground. If you need GPU clusters or colocation space at the end of this year or early next, reach out.