A thread is a pattern that appears at multiple levels of abstraction with the same essential shape. The pattern is introduced at its simplest form — usually a mathematical or algorithmic concept — and then re-appears at each layer of the stack with increasing complexity, different failure modes, and higher stakes.
Thread T5, for example, is Caching. Its root form is memoization: store the result of an expensive function call so you do not recompute it. At the algorithm layer it is dynamic programming — systematic memoization to solve overlapping subproblems. At the infrastructure layer it is a Redis cluster with TTL-based invalidation. At the system design layer it is the CDN in front of a video platform, or the precomputed news feed stored per user. The shape is identical at every layer. The complexity and the consequences of getting it wrong scale upward.