15-minute session: read for 3 minutes, trace for 3, experiment for 7, and complete the check for 2. Spend longer whenever you need.
Build a report for at least two workload sizes. Include source revision, device topology, memory type and placement, affinity, input seed, correctness tolerance, repetitions, raw samples, and end-to-end cost. Distinguish measured results, simulations, and untested candidates. Explain why the winning policy may change on another machine. Submit code and raw evidence on Forge, never confidential machine credentials.
Practice and evidence: Predict the example or draw its state transitions. Change one assumption, test the boundary, and record the actual result. Explain what this small model does not establish about a production system.
Keep three notes: your prediction, the actual result, and why they match or differ. The completion check grades understanding; the practice task builds skill.