Familiarity with liquid-cooled, high-density datacenters and how power, thermal constraints, cooling, and component health affect fleet availability and scheduling and delivery of compute platforms across multiple regions, datacenters, or hybrid-cloud environments, including fleet upgrades and workload migration with minimal service disruption. The fleet spans multiple hardware generations and maturity levels: x86 and Arm servers, GPUs, DPUs, high-speed networking, storage, and rack-scale, liquid-cooled systems.