Ethereum Performance Benchmarking

This repository hosts infrastructure, tooling, and research focused on benchmarking Ethereum client performance under varying gas limits and state sizes. The goal is to enable safe, data-driven protocol tuning and provide actionable insights for the Ethereum ecosystem.

🌐 Project Scope

• Gas Limit Analysis
  Evaluate Ethereum mainnet performance under varying gas limits. Recommend a safe, optimized gas limit based on real data and a public checklist.

• State Bloat Research
  Assess the performance impact of large state sizes (e.g. 10× mainnet) across clients and scenarios. Establish whether state growth meaningfully affects throughput and block processing.

• Infrastructure & Coordination

  • Maintain advanced Perfnets for multiple purposes (stress testing, mainnet-like verification, state bloat perfnet)
  • Provide reproducible benchmarking environments
  • Ensure coordination across Execution and Consensus Layer teams
  • Public dashboards and transparent metrics

📊 Expected Outcomes

• Safe gas limit recommendation for mainnet adoption
• Publicly available benchmarks and performance data
• Clear checklist for evaluating protocol changes
• Streamlined collaboration across clients
• Transparent and trackable process for EL/CL teams

📈 Current Results/Findings

Gas Limit blockers

45 MGas

• OpCodes benchmarking tool having proper execution on performance branches giving enough insgihts for client teams
• Besu Improve EC Scenarios
• General optimizations around ModExp to reach at least 10/12 MGas/s in each EL
• Ensure there is no critical issues on ZKEVM tests

60 MGas

• Further Modexp Improvements for Geth
• Further Modexp Improvements for Erigon
• Point Precompile analysis
• XEN and heavy contracts testing + automation on perfnet
• Full EEST Integration of gas-benchmarks
• Mod/ModArith/Binop_Simple/calldatacopy fixes for Besu
• 2x Bloated state testing

85 Mgas

• Receipts size too big - to be addressed in https://eips.ethereum.org/EIPS/eip-7975
• Modexp Improvements for Besu (if 85 before Fusaka)

100 MGas

• Modexp repricing - https://eips.ethereum.org/EIPS/eip-7883

105 MGas

• Calldata cap - https://eips.ethereum.org/EIPS/eip-7934
• Point Precompile repricing - EIP to be created

State Bloating

• Creating initial infrastructure for State Bloating
• Validating various options for state bloat - Geth/Nethermind plugins or Manual pruning via State-heavy transactions
• Creating first scenarios to bloat a state via Spamoor
• First 50GB bloated
• 1,5X Mainnet Size
• 2X Mainnet Size
• SyncTests executed on 2X state as a baseline comparing to 1X
• 3X Mainnet Size
• 5X Mainnet Size
• 10X Mainnet Size

Various issues reported or fixes made thanks to findings during testing

• [Client_name] Description + Issue Link/PR Link

🧠 Who Is This For?

• Ethereum client developers (EL and CL)
• Core protocol researchers
• Infra engineers and performance analysts
• Ethereum community members interested in network scalability

🤝 Contributions

Contributions are welcome! Please open issues for bugs, ideas, or areas that need clarification.
For major changes or proposals, open a discussion first.

📌 License

MIT License