A shared sequencer is a decentralized sequencing network that provides transaction ordering services to multiple rollups simultaneously, rather than each rollup running its own dedicated sequencer. In the current landscape, Arbitrum’s sequencer orders Arbitrum transactions, Optimism’s sequencer orders Optimism transactions, and Base’s sequencer orders Base transactions — all completely independent. Composability between these chains requires bridges, delays, and trust assumptions. A shared sequencer changes this: by giving a neutral, decentralized network access to transaction ordering across multiple rollups, it enables atomic cross-rollup transactions (swap on Arbitrum and immediately use proceeds on Optimism in a single atomic action), reduces the liveness dependency on individual sequencer operators, and introduces a marketplace model for ordering services. Key projects include Espresso Systems, Astria, and SUAVE (Flashbots).
How It Works
The following sections cover this in detail.
Current Single-Sequencer Model
Each rollup today:
- Has its own sequencer (usually centralized, run by the rollup team)
- Sequences its own transactions
- Posts ordered batches to Ethereum L1
- Is fully independent from other rollup sequencers
Cross-rollup interaction requires: Bridge from Chain A → wait for finality → bridge to Chain B. This takes minutes to hours.
Shared Sequencer Model
With a shared sequencer:
- A decentralized sequencing network (nodes run by many operators) receives transactions for multiple rollups
- The sequencer orders transactions across all subscribed rollups simultaneously
- Cross-rollup transaction bundles can be atomically ordered — the sequencer guarantees “both or neither” across chains
- Ordered transaction sets are posted to L1 (via data availability layer or blobs)
Cross-rollup atomic composability: A user bundles two transactions — “buy ETH on Arbitrum” + “deposit ETH on Optimism” — as an atomic cross-rollup bundle. The shared sequencer includes both or neither, guaranteeing the user doesn’t end up with only one side executed.
Key Shared Sequencer Projects
The following sections cover this in detail.
Espresso Systems
Architecture: An EVM-compatible blockchain (HotShot consensus) that acts as the sequencing layer for subscribing rollups. Rollups opt in to Espresso sequencing, receiving transaction ordering and fast inter-chain finality from Espresso’s consensus.
Key feature: Fast finality (sub-second to a few seconds) with economic security from a staked validator set — enabling composability at sequencer speed, not L1 finality speed.
Astria
Architecture: A decentralized sequencing network using CometBFT consensus, focused on providing a “shared sequencer as a service” that rollups can plugin to. Astria separates sequencing from execution — it only orders transactions, not execute them. This is technically lighter (no EVM needed for Astria nodes).
Key feature: Strong separation of concerns — sequencing, execution, and data availability are separate layers that compose together.
SUAVE (Flashbots)
Architecture: A cross-chain MEV supply chain and ordering network. SUAVE (Single Unified Auction for Value Expression) is designed not just as a sequencer but as a full MEV marketplace — enabling solvers to express complex cross-chain MEV strategies. More oriented toward MEV supply chain than general shared sequencing.
Shared Sequencer vs. Based Rollup
Both shared sequencers and based rollups address the “centralized sequencer” problem in rollups, but via different approaches:
| Feature | Shared Sequencer | Based Rollup |
|---|---|---|
| Who sequences | Decentralized sequencer network | L1 validators (Ethereum itself) |
| Cross-rollup atomicity | Native (shared ordering) | Not native (each rollup L1-sequenced independently) |
| L1 liveness inheritance | Partial (sequencer must be live) | Full (inherits Ethereum L1 liveness) |
| Throughput optimization | High (specialized) | Lower |
| Trust assumption | Sequencer network’s validators | Ethereum validators |
| Maturity | Testnet/early mainnet | Production (Taiko) |
Based rollups inherit L1 security but don’t natively support cross-rollup atomicity. Shared sequencers enable atomicity but introduce a new trust dependency on the sequencer network.
History
- 2022 — Cross-rollup composability identified as a major unsolved problem in Ethereum scaling: siloed L2s can’t compose atomically, limiting DeFi composability across chains.
- 2022–2023 — Espresso Systems founded, backed by major crypto VCs; publishes early shared sequencer architecture.
- 2022–2023 — Astria founded; CometBFT-based sequencing network designed.
- 2023 — Flashbots publishes SUAVE whitepaper, positioning it as a cross-chain MEV and sequencing framework.
- 2023 — “Sequencer wars” — Ethereum community debates based rollups vs. shared sequencers vs. decentralized rollup-specific sequencers as the preferred path forward.
- 2024 — Espresso Sequencer testnet launches; several rollup teams (including based on OP Stack) test Espresso integration.
- 2024–2025 — Astria mainnet launches; first rollups using Astria’s shared sequencing in production.
Common Misconceptions
“Shared sequencers solve all cross-chain problems.”
Shared sequencers solve ordering across chains — they enable atomic transaction inclusion in multiple rollups simultaneously. But final settlement of each rollup still happens independently on L1. Shared sequencing doesn’t eliminate bridge risk for asset movement; it eliminates ordering risk for action coordination.
“Every rollup will use a shared sequencer.”
Rollup teams have incentives to retain their own sequencer: MEV revenue, control over upgrade paths, and independence from external dependencies. Large, established rollups (Arbitrum, Optimism, Base) may never adopt shared sequencers if the benefits don’t outweigh losing sequencer revenue.
“Shared sequencer = shared security.”
Shared sequencing means shared ordering. Security (fraud proof validity, ZK proof correctness) remains per-rollup. “Shared security” in the Ethereum context usually refers to EigenLayer restaking or L1-based security, not sequencing.
Criticisms
- New centralization risk — A shared sequencer that becomes widely adopted becomes itself a critical centralization point. If the shared sequencer goes down or censors, all subscribed rollups are affected simultaneously — worse than individual sequencer failures.
- MEV internalization — Shared sequencers, by ordering across chains, create new MEV opportunities (cross-chain sandwich attacks, timing games). This MEV accrues to the sequencer — whether this is distributed fairly or captured by insiders is an open governance question.
- Limited adoption so far — As of 2025, major rollups (Arbitrum, Optimism, Base) operate their own sequencers and have shown little interest in migrating. Shared sequencers mainly appeal to new rollup launches without established sequencer infrastructure.
- Interoperability standards — Multiple competing shared sequencer protocols (Espresso, Astria, SUAVE) use different interfaces and consensus mechanisms. Rollup developers choosing a shared sequencer must commit to one, risking lock-in.
Social Media Sentiment
Shared sequencers are a DeFi research Twitter/X topic rather than a retail narrative. The “sequencer wars” discourse of 2023 generated significant niche engagement, with Ethereum researchers debating the merits of shared vs. based vs. decentralized-but-rollup-specific sequencing. Rollup ecosystem accounts (Optimism, Arbitrum Foundation) rarely engage with shared sequencer proposals publicly. Espresso and Astria have active respective communities. The topic is seen as important infrastructure-in-progress — technically sophisticated, potentially transformative for cross-chain composability, but not yet producing consumer-visible results.
Last updated: 2026-04
Related Terms
Sources
- Espresso Systems (2023). HotShot Consensus and the Espresso Sequencer. espressosys.com/whitepaper.
- Astria (2023). Astria: A Shared Sequencer Network. astria.org/docs.
- Flashbots (2023). SUAVE: Towards a Decentralized and Private MEV Supply Chain. blog.flashbots.net.
- Drake, J. (2023). Based Rollups: Superpowers from L1 Sequencing (Shared Sequencer Comparison Section). ethresear.ch.