AI capacity is rarely in one place anymore. Operators run GPU fleets across several regions, each with its own hardware generation, power budget, cooling design, and cost. When every site is managed on its own, GPUs sit idle in one data center while jobs wait in a queue at another.
AboveCloud is part of Federator.ai Cortex, ProphetStor’s full-stack AI Ops platform for AI factories. It extends Cortex from a single site to an entire fleet, giving operators one control plane to see every data center, compare them, and place or move workloads where they run best.
What AboveCloud Delivers
Federated Global View
Every data center in the fleet appears on one live map, with capacity, GPU inventory, PUE, renewable share, and inter-site latency side by side. Operators compare sites at a glance instead of switching between consoles.
ERIC Scoring-Based Site Ranking
AboveCloud scores each data center on Energy, Reliability, Interconnect, and Compute, then ranks the whole fleet. The ranking shows where new workloads will run best and which sites can absorb more demand.
Fleet-Wide Placement and Migration
Workloads, including large LLM training runs, are submitted once and placed in the best-scoring data center. If conditions change, kMotion moves a running job to another site without restarting it from scratch.
A Digital Twin of Every Site
Each data center has a digital twin, from the floor plan down to individual racks. Heat map and coolant flow views let operators trace a fleet-level alert down to the exact rack that is causing it.
What-If Financial Planning
Before committing capacity or moving workloads, teams can model scenarios and compare projected cost, savings, and return on investment. Finance and operations plan from the same numbers instead of separate spreadsheets.
Toward a Global AI Compute Marketplace
AboveCloud platform is designed to go beyond a single operator’s fleet. Its marketplace framework lets data centers offer spare GPU capacity and match it to demand through reverse auctions and spot pricing, so idle GPUs in one region can serve workloads from another.
Run Many Data Centers as One
Every new site adds capacity, but it also adds complexity. AboveCloud keeps a growing fleet manageable by turning separate data centers into a single, coordinated resource, so each GPU, wherever it sits, contributes to the work that matters most.
Frequently Asked Questions
What is the AboveCloud platform?
AboveCloud is a Federator.ai Cortex solution for orchestrating AI workloads across multiple data centers. It brings every site in a GPU fleet into one control plane, so operators can compare data centers, place workloads, and move them between sites from a single view.
How does AboveCloud decide where a workload should run?
AboveCloud ranks every data center using ERIC, ProphetStor’s scoring framework that weighs Energy, Reliability, Interconnect, and Compute. New workloads go to the site that offers the best balance for their needs, and the ranking updates as conditions change.
Can AboveCloud move a workload to another data center after it starts?
Yes. AboveCloud uses kMotion, another Federator.ai Cortex solution, to live-migrate running jobs between data centers. The job resumes at the new site without restarting from scratch.
What does the AboveCloud digital twin show?
The AboveCloud digital twin models each data center from the floor plan down to individual racks, with heat map and coolant flow views. Operators can move from a fleet-wide view to a single rack in a few clicks to find the source of a thermal or power issue.
Can data centers trade GPU capacity through AboveCloud?
AboveCloud is designed to support a global AI compute marketplace, where data centers offer spare GPU capacity and match it to demand through reverse auctions and spot pricing.