Layer 2

How L2s work

The basic idea: process transactions off the main chain (L1), then post a compact summary back to L1. The off-chain processing is fast and cheap; the on-chain summary inherits L1's security guarantees.

A typical L2:

1. Users submit transactions to the L2.
2. The L2's sequencer orders and executes them quickly.
3. The L2 batches many transactions and posts compressed data to L1.
4. Either via fraud proofs (optimistic rollups) or validity proofs (ZK-rollups), L1 verifies that the L2's state changes are valid.
5. Once finalized on L1, transactions are as secure as L1 itself.

Result: users get fast, cheap transactions while inheriting Ethereum's security.

Major L2 categories

Two dominant types:

• Optimistic rollups — assume validity by default; allow challenges via fraud proofs during a 7-day window. Arbitrum, Optimism, Base are the largest examples.
• ZK-rollups — use cryptographic validity proofs to demonstrate state changes are valid. zkSync, Starknet, Polygon zkEVM, Linea, Scroll are examples.

Other approaches (state channels, plasma, validiums) exist but have less traction than rollups.

Major L2s

A few worth knowing:

• Arbitrum — largest L2 by TVL. Optimistic rollup. ARB token, mature DeFi ecosystem.
• Optimism — second-largest. Optimistic rollup. OP token, OP Stack widely adopted.
• Base — Coinbase-built; uses OP Stack. No token. Largest by daily transactions.
• zkSync Era — major ZK-rollup; ZK token launched 2024.
• Starknet — Cairo VM-based ZK rollup; uses STRK token.
• Polygon zkEVM, Polygon PoS — different scaling approaches under the Polygon brand.
• Linea — Consensys-built ZK rollup.
• Scroll — community-driven ZK rollup.
• Mantle — combination L2 with custom data availability.
• Blast, World Chain, others — newer L2s with specific focuses.

Why L2s matter

The economic logic:

• Ethereum mainnet is expensive ($5-50 per transaction in normal conditions, $100+ during congestion).
• L2s are 10-100x cheaper for typical transactions.
• Most user activity can happen on L2 with sufficient security for the use case.
• L1 stays as a settlement layer for high-value or final operations.

This "rollup-centric roadmap" is Ethereum's official scaling strategy. Total L2 transaction volume now exceeds Ethereum mainnet by significant multiples.

EIP-4844 and blob fees

A pivotal upgrade in March 2024:

• Pre-4844 — L2s posted data to Ethereum as expensive calldata.
• Post-4844 — L2s use cheap "blobs" — temporary data availability with shorter retention.

The result: L2 transaction costs dropped 90%+ overnight. This made L2s economically viable for previously-uneconomic use cases (gaming, microtransactions, social).

L2 limitations

Several real concerns:

• Centralized sequencers. Most L2s currently have a single sequencer that orders transactions. Single point of failure or censorship.
• Withdrawal delays. Optimistic rollup withdrawals take 7 days; users typically use third-party "fast bridges" with their own trust assumptions.
• Cross-chain UX. Moving between L2s (and between L2s and L1) involves bridges with their own complexity.
• Fragmentation. Liquidity, applications, and users are split across many L2s.
• Validator economics. L2 economics differ from L1; long-run sustainability of various L2 models is being tested.

OP Stack and the Superchain vision

Optimism's OP Stack has become the dominant framework for new L2s:

• Base — Coinbase-built using OP Stack.
• World Chain (Worldcoin) — uses OP Stack.
• Mode, Zora, others — many smaller L2s use OP Stack.

The "Superchain" vision is that OP Stack chains share security and infrastructure, enabling atomic interoperability between them. This isn't fully realized yet but represents a major direction.

Specialized L2s

Some L2s target specific use cases:

• Blast — yield-focused; native ETH and stablecoins earn yield.
• World Chain — Worldcoin-specific; humans-only chain.
• Various gaming-focused L2s — optimized for high-throughput, low-fee gaming use cases.
• Restaking L2s — using EigenLayer for security.

Specialization is increasing as the L2 space matures.

L1 vs. L2 economics

A meaningful tension:

• L2s capture transaction-fee revenue that previously went to L1 validators.
• L1 (Ethereum) keeps some revenue through L2 data costs and base settlement.
• The split affects which token captures more value: ETH or L2 tokens like ARB, OP.
• Long-term ETH economics depend partly on whether L1 fees can sustain validator rewards as activity moves to L2.

This dynamic is one reason ETH price action has sometimes diverged from broader Ethereum-ecosystem activity. The activity moves to L2s; the value capture is contested.

What individuals should know

For users:

• Most active users now operate primarily on L2s rather than Ethereum mainnet.
• Major L2s are reasonably mature for most use cases.
• Cross-L2 movement requires bridges with their own trade-offs.
• Costs and speed are dramatically better than mainnet for typical operations.

For investors in tokens:

• L2 tokens have variable economic models. Some accrue value through fees; others mostly through governance.
• No token L2s like Base don't have a token to invest in directly; exposure is through Coinbase or similar.
• L1 tokens like ETH retain value through their settlement role even as activity moves to L2.

The L2 ecosystem represents the most-active scaling work in crypto. Whether the L2-centric model wins, or alternative L1 architectures (Solana, Monad) regain relative importance, is one of the more consequential open questions in the space.