Vitalik Buterin: Future of Ethereum Scaling & L1 Considerations

The Need for L1 Scaling

• 💡 The speaker argues for higher L1 gas limits even in a world dominated by Layer 2 (L2) solutions, emphasizing the continued value of the L1.
• 🎯 A core goal is to ensure every Ethereum user can occasionally use the L1, aiming for a capacity that supports as many users as Facebook (3.1 billion).
• 📈 Achieving this user base would require the L1 to handle 100 transactions per second, necessitating approximately a 6x increase in current capacity.

Critical L1 Use Cases & Costs

• 🔒 Censorship resistance is vital for DeFi protocols and L2 force inclusion mechanisms, but current L1 costs mean a bypass operation is around $4.50, requiring a 4.5x L1 scale-up for affordability.
• 🌉 Moving low-value assets and NFTs between L2s often necessitates direct L1 transactions, which are currently expensive and would demand a 6x capacity expansion for mass adoption.
• ⚠️ L2 mass exits to L1 must be affordable and feasible for all users within a week, with current calculations suggesting the L1 needs to scale by a factor of 1 to 9 times.
• 🪙 Issuing ERC20 tokens on L1 is considered safer against hostile L2 governance, but current costs are high, potentially requiring an 18x L1 scaling for frequent market creation.

Optimizing L1 for L2 Efficiency

• 🔑 Key store wallets, which store wallet data on L1, can be optimized to a lower gas cost (around 7500 gas), bringing them closer to a target cost of 28 cents.
• ⚡ L2 proof submission is crucial for fast, trustless interoperability between L2s, but current costs are prohibitive ($50 million/year), highlighting the need for aggregation protocols to reduce costs significantly (to $100K/year).

Defining L1 Decentralization Goals

• 🧠 It's essential to rigorously define the underlying goals for L1, including acceptable disk, compute, and bandwidth requirements for different node types (verifier, staker, builder).
• ⚖️ Implementing stateless clients can reduce storage for verifiers but potentially increase it for block builders if gas limits rise, necessitating careful consideration of decentralization trade-offs.
• 🌐 The speaker argues that compute requirements are more robust against internet censorship than bandwidth requirements, suggesting a strategic shift towards prioritizing compute in L1 optimizations for long-term resilience.