From High Fees to High Speeds: Optimizing the SVM For Ethereum
summary
Eclipse claims to deliver an ultra‑fast L2 by forking the Solana VM and changing it to produce verifiable execution that settles to Ethereum with fraud proofs and later ZK proofs, while using Celestia for data availability. The design accepts a small, very expensive sequencer set to run beefy hardware, trades decentralization for provable correctness, targets throughput beyond current EVM L2s (counterpoint: Coinbase Base aims ~2.5M gas/sec), and notes ZK proof overheads of ~100k–1M× per execution.
AI-written summary · about
chapters (12)
- 0:00Eclipse introduction
- 0:54Layer twos performance
- 2:20Data availability bottlenecks
- 3:44Parallelism is not the most important optimization
- 4:48Optimizing L2 first principles
- 5:31Access lists advantages
- 6:32Northstar performance
- 7:55L1 vs L2 verifiability
- 9:21Sequencer superpower
- 10:44ZK rollup upgrades
- 11:27Thick sequencer model
- 12:10Closing remarks
related talks
09mEclipse: Ethereum's First SVM L2
19mAccelerating Ethereum: Fast Finality Rollups and the Movement Ecosystem
26mThe MEV Game on L2: Bots, Users, Sequencers Dynamics
29mEthereum Sharing
09mReflections on Solana From an EVM Builder
01h 10mMEVconomic's Panel: Speedrunning Finance Latency in MEV
26mWhat's New in MEV
25mZK Rollup: Ethereum Scalability with ZKPs