Skip to content

Layer 2

Layer 2 is Ethereum's answer to the same scaling tension Bitcoin's Bitcoin Scaling section covered: keep the base chain deliberately constrained so running a full node stays accessible, and push transaction throughput onto separate systems that inherit L1's security through specific, verifiable mechanisms rather than a fresh trust assumption. This section covers those mechanisms directly, including several dated, sourced facts about where the major rollups actually stand today, not just how they're designed to work eventually.

What you need to know first

Gas, Proof of Stake, and Zero-Knowledge Proofs. This section assumes familiarity with Ethereum's account and gas model, and builds directly on the general zero-knowledge proof concepts already covered in cryptography.

Chapters

  1. L1 vs. L2: why Ethereum didn't just raise L1's own throughput instead
  2. Rollups: the sequence-batch-verify structure every rollup design shares
  3. Optimistic Rollups: why withdrawals take about a week, worked through directly
  4. Fraud Proofs: interactive bisection, narrowing a dispute to one cheap-to-verify step
  5. ZK Rollups: proof instead of assumption, and the real cost of proof generation
  6. Validity Proofs: what statement a proof actually establishes, and what a zkEVM circuit has to reproduce
  7. Sequencers: the current, dated fact that every major rollup's sequencer is centralized
  8. Data Availability: why availability and correctness are separately enforced guarantees
  9. Blobs: Ethereum's dedicated, cheap, temporarily-retained data type for rollup batches
  10. EIP-4844: the blob base fee formula, derived and verified with real numbers
  11. Bridges: the trust spectrum from federated multisigs to trust-minimized light clients
  12. Canonical Bridges: why a rollup's own bridge reuses its existing verification mechanism
  13. Cross-Chain Messaging: LayerZero's configurable verifiers and Wormhole's 19-guardian network
  14. Arbitrum: Nitro, and the real gap between theoretical and actually-deployed permissionless fraud proofs
  15. Optimism: the OP Stack, the Superchain vision, and where that vision actually stands
  16. Base: Coinbase's L2, and a dated, sourced example of measuring decentralization directly
  17. zkSync: EVM-compatible but not EVM-equivalent, with native account abstraction
  18. StarkNet: Cairo, STARKs specifically, and an independently designed account model

Next

Continue to Security, where the trust boundaries this section has drawn carefully (what a sequencer can and can't do, what a bridge's specific verification mechanism actually guarantees) become the starting point for understanding how real exploits have broken those boundaries in practice.