Your question is GPU Job Scheduling. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How would you schedule jobs on a GPU fleet?
Asked in the final round stage. Cover scheduling primitives, model sizes, cold start times, traffic prioritization, and queue management. Reported follow-ups include model sizes and cold start times.
Explain the control plane, admission and placement decisions, queueing policy, preemption behavior, and how you would measure the system. Address both latency-sensitive inference and batch or distributed training workloads, while calling out the information you would request before sizing the fleet.