One number. Same everywhere.

The graph-traversal headline used to read one way on the scorecard, another way through the reproducibility API, and a third way on the live demo, because each was measuring a slightly different thing at a different moment. It now comes from a single trailing source with a single method label, so the scorecard, the API and the demo all read the same figure.

graph traversal · the record + its neighbors, over the live value-flow

493.2MTEPS

warm warm · in-process · single-shard · all-cores measured ● LIVE

warm working-set BFS over the resident value-flow graph, the trailing average across live runs

the scorecard

493.2M TEPS warm

warm · in-process · single-shard · all-cores measured

The overview row reads this figure and this tier, live on this box.

open the scorecard ▸

the reproducibility API

493.2M TEPS warm

warm · in-process · single-shard · all-cores measured

GET /api/bench/graph_teps?dataset=btc-chain runs the same bounded traversal and returns the same class of number, measured live when you call it.

the live demo

493.2M TEPS warm

warm · in-process · single-shard · all-cores measured

The graph demo runs it in front of you and carries this same method label.

run the demo ▸

The figure is a warm trailing average across live runs, tier-labeled honestly. A single run you trigger yourself will land near it, not exactly on it, because it is one sample and this is the moving average, that is what a live, reproducible number looks like. watch the whole suite run ▸

Seen enough? Every number here is measured live on public data, and moves every week. Talk to the team ▸