Experiment B · Network entanglement · shared-bath topology race
13 cavities. 28 edges.
Which topology survives the noise?
Ring, line, hub, complete, atlas, and the Fibonacci-shell topology derived from the Atlas v3 result on `/atlas`. Same N. Same single-node Q. Same bath. The bath correlation between two cavities decays exponentially with their separation. We compute the per-edge concurrence under shared dephasing and aggregate.
View
Topology picker + race result + visualisation. The minimum to read the experiment.
EXPLAIN LIKE I'M 10
· everyday-language one-paragraph version
Imagine twelve quantum stations spread across a country, all sharing the same weather. The weather wrecks a quantum link, but if two stations share the SAME weather, the link survives. So the question is: should the stations be wired in a ring? A line? A star? All-to-all? Each shape pays a different price for connection. The page races six shapes against each other under the same weather and tells you which one keeps the most links alive, including the new Fibonacci-shell shape derived from the Atlas v3 finding.
WHAT YOU ARE LOOKING AT
A network of N quantum cavities, each at one geographic point, all subject to shared environmental noise whose correlation falls off with distance. Each edge of the graph is a possible quantum link. The page computes the bipartite entanglement preserved on every edge after a fixed operating time, then aggregates to a network total and a per-node average.
WHY THIS MATTERS FOR QUANTUM
Distributed quantum tasks (entanglement-based key distribution, distributed sensing, multi-node clock synchronisation) depend on the network preserving entanglement long enough to be useful. Different topologies pay different costs: a complete graph has the most edges but every extra edge is a new pathway for decoherence. A ring has the fewest. The question is which topology gives the best entanglement-per-edge × edges-per-node trade-off at a given environmental coherence length.
WHAT TO WATCH
Bath coherence length L_coh sets how correlated noise is between neighbouring nodes. Short L_coh = noise is local = correlated decoherence-free subspace fails. Long L_coh = nearby cavities share their bath = entanglement persists. The race table shows each topology's E_avg under your current parameters. The atlas topology uses real Jyotirlinga coordinates, so the bath correlations are physically meaningful, not synthetic.
Topology
atlas
13 nodes · 28 edges
E_net (total)
25.962
sum of per-edge concurrences
E_avg (per node)
1.997
E_net / N
Best topology @ same params
complete
E_avg=3.524
Network visualisation · edge brightness = concurrence
orange = high E · dim = decohered
Topology race · same parameters
| Topology | Nodes | Edges | E_net | E_avg | Verdict |
|---|---|---|---|---|---|
| complete | 12 | 66 | 42.294 | 3.524 | strong |
| atlas | 13 | 28 | 25.962 | 1.997 | strong |
| fibonacci | 12 | 26 | 23.017 | 1.918 | strong |
| ring | 12 | 12 | 8.750 | 0.729 | strong |
| line | 12 | 11 | 8.027 | 0.669 | strong |
| hub | 12 | 11 | 7.032 | 0.586 | strong |
Best topology depends on bath correlation length L_coh. Long L_coh favours dense topologies (more edges = more entanglement); short L_coh favours sparse ones (each extra edge is a new local-noise channel). The atlas topology bakes in real Jyotirlinga separations.