RESPONSIBILITIES: Partner deeply with responsible engineers, thermal engineers, and server mechanical teams to co-define commodity strategy, multi-year technology roadmaps, and approved vendor lists for server mechanics and cooling - chassis, lids, bezels and GPU/accelerator trays; cold plates (CPU, GPU, HBM, NIC, switch ASIC); heat sinks, vapor chambers and heat pipes; thermal interface materials (grease, pads, PCM); server fans and shrouds; node-level liquid loops (tubing, fittings, on-server manifolds, dripless QDs); and retention hardware (GPU brackets, CPU ILM/backplates, DIMM clips, NIC brackets) - anticipating SpaceXAI TDP and density requirements 12-24 months ahead. Own global sourcing end-to-end for server mechanicals and cooling, including supplier selection, qualification of die-cast / stamped chassis, cold-plate, TIM, and fan suppliers, dual/multi-sourcing on constrained thermal SKUs, capacity reservation, and performance governance to achieve on-time delivery and continuous lead-time compression while supporting SpaceX's aggressive cluster build cadence.