Scaling & Layer 2 Solutions
Chapter 8 closed on a real, concrete demonstration of a limit: CryptoKitties alone, at its peak, consumed up to 70% of Ethereum's entire network capacity. This chapter explains precisely why a base blockchain has so little headroom to spare in the first place, and covers the real, working approaches the industry has built to scale far beyond what a single base chain can process alone.
The Scalability Trilemma
Blockchain designers face a genuine, well-known tension often described as the scalability trilemma: a blockchain design tends to be able to strongly deliver on only two of these three properties at once, without real, significant compromise on the third:
many independent participants, no single point of control
genuinely resistant to attack (Chapter 4's own 51%-attack economics)
handling a large volume of transactions quickly and cheaply
Bitcoin and Ethereum's own base layers, as covered so far in this course, both deliberately prioritize decentralization and security. That choice comes with a real, direct scalability cost.
The Real Throughput Gap
Bitcoin's own real, documented base-layer throughput sits at roughly 7 transactions per second — a direct, structural consequence of Chapter 3's own fixed block size and Chapter 4's own roughly-10-minute block interval. Visa's own published figures describe its data centers as capable of handling up to 30,000 simultaneous transactions — a real, order-of-magnitude reminder of just how much headroom a base blockchain layer alone doesn't have if it needs to compete with existing centralized payment infrastructure at real-world scale.
Layer 1 vs. Layer 2
Layer 1 refers to the base blockchain itself — Bitcoin's or Ethereum's own chain, with all the security properties this course has built up over the last eight chapters. Layer 2 refers to a separate system built on top of a Layer 1 chain, designed to handle the bulk of everyday transaction volume off the base chain, while still ultimately relying on the Layer 1 chain underneath for its own final security guarantees.
Payment Channels: Bitcoin's Lightning Network
Bitcoin's real, working Layer 2 solution is the Lightning Network, first described in a whitepaper by Joseph Poon and Thaddeus Dryja in February 2015, and actually launched on Bitcoin's mainnet in 2018. It works through payment channels: two parties lock up funds together in a shared, multi-signature on-chain transaction, then conduct any number of instant, effectively free transactions between themselves entirely off-chain, simply updating the private, agreed-upon split of the locked funds each time. Only when the channel is finally closed does one final, settled transaction get broadcast to the actual Bitcoin blockchain.
A network of many interconnected channels lets payments route between people who've never directly opened a channel with each other, hopping across a path of existing channels. This real network has grown substantially: as of recent measurements, it spans roughly 17,000 public nodes and around 40,000 public channels, with a total real capacity of roughly 4,900 BTC locked up across the network.
Rollups: Ethereum's Primary Scaling Approach
Ethereum's own dominant real scaling strategy is built around rollups, which execute a large batch of transactions off the main Ethereum chain, then post a compact summary of the results back to Layer 1, inheriting Ethereum's own base-layer security for that summary. Two real, genuinely different approaches exist:
- Optimistic rollups — assume every batch of transactions is valid by default, and only actually run the underlying computation to verify it if someone specifically challenges the result during a set dispute window. This keeps everyday operation cheap, at the real cost of a withdrawal delay while that challenge window remains open.
- Zero-knowledge (ZK) rollups — compute the results off-chain and submit a real cryptographic proof, verified mathematically on Layer 1, that the batch was processed correctly — without needing to reveal or re-execute every individual transaction on the base chain. This avoids the optimistic rollup's own withdrawal-delay trade-off, at the real cost of more complex cryptography to implement correctly.
| Approach | Real Example | Real, Honest Trade-off |
|---|---|---|
| Payment channels | Lightning Network (Bitcoin, 2018) | Requires locking up capital in advance and actively managing channels |
| Optimistic rollups | Ethereum ecosystem rollups | Cheap, simple; withdrawals delayed by a dispute/challenge window |
| ZK rollups | Ethereum ecosystem rollups | Fast finality, no challenge window; more complex cryptography to build correctly |
| Sidechains | Various bridged chains | Own independent security model — doesn't inherit Layer 1 security automatically |
Hands-On Exercises
Three exercises reinforcing the scalability trilemma and real Layer 2 mechanics, before Chapter 10's capstone traces one real transaction all the way through everything this course has covered.
Quick Reference
- Scalability trilemma — decentralization, security, and scalability are genuinely hard to maximize all three at once.
- Real throughput — Bitcoin's base layer processes roughly 7 transactions/second; Visa's own data centers report handling up to 30,000 simultaneously.
- Layer 1 vs. Layer 2 — the base chain itself vs. a system built on top of it to offload everyday transaction volume.
- Lightning Network — Bitcoin's real payment-channel Layer 2, launched 2018, letting parties transact off-chain and settle only occasionally on-chain.
- Optimistic rollups — assume validity by default, verified only if challenged within a dispute window; cheap, but delayed withdrawals.
- ZK rollups — submit a cryptographic proof of correct execution; fast finality, but more complex to build.
- Sidechains — separate, independently-secured chains bridged to Layer 1, not automatically inheriting its security.