Throughput / Latency
Kind: tension
Record: tension:latency-throughput
Canonical: Schema
"throughput" (performance-core layer) is traded against "Latency" (performance-core layer) — a principle cannot be scope-separated from a quality, metric, or cost it competes with; resolve by measuring "Latency" and choosing an explicit operating point.
Listed in Tensions, after Parallelism / Coordination Overhead and before Latency / Throughput/Batching.