Performance also: USL

Universal Scalability Law

C(N) = N / (1 + α(N−1) + βN(N−1)). Throughput vs concurrency.


In plain terms

Above a critical N, adding workers slows throughput. β captures coordination cost (O(N²)). The crisper version of "Amdahl plus contention plus coherence".

Origin

Neil Gunther formulated USL in the 1990s, building on Amdahl with an extra term for coordination cost. Predicts the negative-return region — beyond a critical N, more workers makes things slower.

Where it shows up in production
  • Database lock contention At some concurrency threshold, throughput inverts and falls. USL fits the curve.
  • Scalability modeling Fit α and β to throughput-vs-load data and predict where adding capacity will stop helping.
On Semicolony
Sources & further reading
Found this useful?