LAYER 2

A Technical Wiki

L1 Rollups Channels Bridges App

L1 Consensus: The Bedrock

The base layer is the foundation on which everything rests. Every node validates every transaction. Consensus is global, final, and expensive. The design is intentionally conservative -- security above all, throughput as a secondary concern.

Layer 1 processes roughly 15 transactions per second. This is not a limitation to be fixed at the base layer. It is a design constraint that Layer 2 solutions respect and work around. The bedrock does not need to be fast. It needs to be unbreakable.

Finality: ~12 minutes (Ethereum PoS) | Block time: 12s | Gas limit: 30M

CONSENSUS LAYER Validator nodes

Rollups: Compression at Scale

Rollups execute transactions off-chain, then post compressed transaction data back to Layer 1. The name is literal: hundreds of transactions are rolled up into a single proof submitted to the base chain.

Two paradigms dominate. Optimistic rollups assume all transactions are valid and allow a challenge window (typically 7 days) for fraud proofs. ZK-rollups generate cryptographic validity proofs that mathematically guarantee correctness before the batch is accepted.

Compression ratio: ~10-100x | Finality: seconds (L2) to minutes (L1) | Data availability: on-chain

L1 Optimistic ZK Fraud proofs Validity proofs
ROLLUP COMPRESSION Raw transactions 1000 transactions compress Batch ~10KB calldata prove L1 Post 1 tx on-chain

State Channels: Off-Chain Dialogue

State channels allow two or more parties to transact thousands of times off-chain, with only the opening and closing states recorded on Layer 1. The channel is literally a channel -- a private conduit between participants.

The elegance lies in the economics: n transactions cost roughly the same as 2 on-chain transactions (open + close). Payment channels, gaming channels, and communication channels all share this fundamental structure.

On-chain footprint: 2 transactions | Throughput: unlimited (between parties) | Finality: instant

Alice Bob n transactions L1: open + close only
STATE CHANNEL HANDSHAKE Party A Party B Open Transact x N Close L1 settlement: 2 on-chain transactions total

Bridges: Interoperability Infrastructure

Bridges connect Layer 2 networks to each other and back to Layer 1. They are the interoperability layer -- the mechanism by which assets and messages cross between isolated execution environments.

The security model of a bridge is its defining characteristic. Trust-minimized bridges rely on cryptographic proofs or fraud-proof mechanisms. Trusted bridges depend on a set of validators or a multisig. The bridge trilemma -- trustlessness, extensibility, generalizability -- defines the design space.

Security model: varies (trusted to trust-minimized) | Latency: minutes to days | Risk: bridge exploits represent the largest category of DeFi losses

L2-A L2-B B L1 Settlement
BRIDGE SECURITY MODEL Trust-Minimized ZK proofs / fraud proofs Higher latency Maximum security Optimistic Challenge period Moderate latency Good security Trusted Multisig / validators Low latency Trust assumption

Application Layer: Where Users Live

The application layer is where Layer 2 scaling becomes invisible. Users interact with dApps, swap tokens, mint assets, and play games without knowing which rollup, channel, or sidechain processes their transactions.

This is the goal of Layer 2: to make the infrastructure disappear. The best scaling solution is the one you never notice. When a transaction costs $0.001 and confirms in 2 seconds, the user is simply using the internet. The layers beneath are architecture, not experience.

The best infrastructure is invisible infrastructure.

User Interface L2 execution (invisible) L1 settlement (invisible)