Shared Sequencing
Shared sequencing lets multiple rollups use one decentralized sequencer network instead of each running its own centralized one — Espresso and Astria pione
Shared sequencing lets multiple rollups use one decentralized sequencer network instead of each running its own centralized one — Espresso and Astria pioneered the model. One neutral ordering layer serving many chains, with atomic cross-rollup inclusion as the killer feature.
How it works
Rollups post transactions to the shared network, which orders them (across chains, atomically) and returns sequences each rollup executes. Ordering power moves from single operators to a staked, rotating validator set with its own slashing and governance.
Why it matters for decentralization
It attacks the single biggest rollup centralization vector directly — and replaces N trusted sequencers with one shared trust assumption to audit hard. Genuinely decentralized shared sequencing would upgrade the whole L2 ecosystem at once; a captured one would compromise it at once. Same leverage, both directions.
Risks & trade-offs
New trust layer (who runs it?); cross-chain MEV complexities; sovereignty questions (whose ordering rules win disputes?); and liveness coupling — shared downtime hits every member simultaneously.
FAQ
Shared vs based sequencing? Shared = neutral third-party network orders for many rollups. Based = L1 validators do the ordering (maximal L1 alignment, minimal new trust). Different bets, same problem.
Does it fix MEV? It relocates it to the shared layer — which must then solve fair ordering itself. Turtles all the way down, but shared turtles.
Live yet? Testnets and early integrations; production shared sequencing at scale is still ahead. Track operator sets, not announcements.