Layer 2 (L2) solutions are a cornerstone of Ethereum's scaling strategy, offering users lower gas fees and faster transaction throughput while inheriting the security of the Ethereum mainnet. This article provides a comprehensive analysis of the current L2 landscape, focusing specifically on two critical performance metrics: transaction fees and maximum achievable transactions per second (TPS).
Understanding Layer 2 Growth and Importance
The L2 ecosystem has experienced remarkable growth since the launch of the first project in late 2019. Recent data indicates that over 60% of all Ethereum transactions now occur on L2 networks, demonstrating their increasing adoption and importance. This growth represents a scaling factor of approximately 5.49x compared to Ethereum's base layer capacity.
Various approaches have emerged to address Ethereum's scalability challenges, with development now centered primarily on rollup technologies. These solutions aim to maintain Ethereum's security guarantees while significantly improving performance metrics that directly impact user experience.
Key Performance Metrics Explained
When evaluating L2 solutions, two factors primarily determine practical usability:
Transaction Fees: The cost users pay to execute operations on the network
Maximum TPS: The theoretical upper limit of transactions the network can process per second
These metrics vary significantly across different L2 implementations due to their architectural differences and data availability approaches.
Comprehensive Fee Structures Across L2 Solutions
Transaction fees on rollup networks typically consist of two main components:
Execution Fee
This covers the computational work required to process transactions on the L2 network. It follows a structure similar to Ethereum's gas model, calculated as:
execution_fee = gas_used × gas_priceData Availability Fee (Rollup Fee)
This covers the cost of storing transaction data on Ethereum's base layer, ensuring security and verifiability. This component often represents the significant portion of total transaction costs.
Different L2 implementations employ distinct fee calculation methods:
OP Stack and OVM-Based Solutions
Projects like Optimism, Base, and Kroma use a transparent fee model:
total_fee = l2_execution_fee + l1_data_feeThe L1 data fee calculation incorporates fixed and dynamic overhead parameters that vary between implementations.
Arbitrum's Approach
Arbitrum uses a unique formula that includes an extra buffer for L1 costs:
transaction_fee = l2_gas_price × (l2_gas_used + (l1_calldata_price × l1_calldata_size) / l2_gas_price)This approach accounts for the brotli compression algorithm used for batch data.
Zero-Knowledge Rollup Variations
ZK rollups incorporate additional costs for proof generation and verification:
Starknet calculates computational complexity fees based on Cairo VM operations
zkSync Era employs a multi-factor pricing model including proof generation costs
Scroll uses a hybrid model similar to OP Stack but with different parameter configurations
👉 Explore real-time fee comparison tools
Transaction Fee Comparison Analysis
Based on comprehensive research across 15 major L2 projects, we observe clear patterns in fee structures:
- Optimium solutions (like Mantle) generally offer the lowest fees as they don't publish data on-chain
- Optimistic rollups typically provide mid-range pricing
- ZK rollups tend to have higher fees due to proof generation costs
- All L2 solutions offer transaction costs significantly lower than Ethereum—typically by a factor of 10x or more
The specific fee for a transaction type (ETH transfer, ERC-20 transfer, or swap) varies based on each network's implementation details and current network conditions.
Bridge Costs: Deposits and Withdrawals
Bridge operations between L1 and L2 incur separate costs that differ from internal L2 transactions:
Deposit Costs
Depositing assets from Ethereum to L2 involves L1 execution costs, making these operations generally more expensive than internal L2 transactions. Projects based on OP Stack typically offer the most competitive deposit fees.
Withdrawal Processes
Withdrawal mechanisms vary significantly across different L2 solutions:
- OP Stack solutions require a multi-step process with a 7-day challenge period
- Arbitrum uses cross-chain messaging with different cost structures
- ZK rollups often have simpler withdrawal processes with lower costs
- Some solutions require native tokens for withdrawal operations
Maximum Throughput Calculations
The maximum TPS of a rollup is fundamentally limited by how quickly transactions can be verified on Ethereum. We calculate this using:
max_tps = max_transactions_per_batch / min_batch_intervalThroughput Analysis by Technology
OP Stack Solutions: Achieve approximately 455 TPS through efficient compression and frequent batch submission
Arbitrum One: Currently achieves around 226 TPS with potential for improvement
zkSync Era: Leads with 750 TPS due to aggressive batch intervals and efficient proof systems
Starknet: Focuses on proof optimization rather than maximum throughput, achieving approximately 25 TPS
Scroll: Implements a multi-layer batching system resulting in 50 TPS
Optimium Solutions: Projects like Metis and Mantle achieve 350-400 TPS by using alternative data availability layers
These values represent theoretical maximums under ideal conditions. Real-world performance is typically lower due to network congestion, variable transaction sizes, and practical implementation constraints.
The Impact of EIP-4844 and Future Developments
The upcoming Ethereum Dencun upgrade, featuring EIP-4844 (proto-danksharding), promises to revolutionize L2 economics by introducing blob storage. This innovation is expected to:
- Reduce data availability costs by 10-100x
- Enable higher throughput by allowing larger batches
- Improve cost predictability for L2 operations
- Benefit Optimium solutions particularly significantly
Most L2 teams are awaiting this upgrade before implementing additional cost optimization strategies.
Frequently Asked Questions
What determines transaction fees on Layer 2 networks?
Transaction fees consist mainly of execution costs (for processing the transaction on L2) and data availability costs (for storing data on Ethereum). The balance between these components varies across different L2 implementations and is influenced by current network conditions on both layers.
Why do ZK rollups generally have higher fees than optimistic rollups?
ZK rollups incorporate additional costs for generating cryptographic proofs that verify transaction validity. These computational requirements add to the overall transaction cost compared to optimistic approaches that rely on fraud proofs and challenge periods.
How will EIP-4844 affect Layer 2 transaction costs?
EIP-4844 will introduce blob storage to Ethereum, creating a dedicated data space for rollups at significantly reduced costs. This is expected to lower L2 transaction fees substantially, particularly for operations that require extensive data availability.
What is the relationship between batch intervals and transaction throughput?
Shorter batch intervals allow more frequent submission of transactions to Ethereum, increasing potential throughput. However, practical limitations include Ethereum block times, network congestion, and the computational requirements for preparing and compressing batches.
Are advertised TPS figures realistic for everyday usage?
Most published TPS figures represent theoretical maximums under ideal conditions. Real-world performance typically falls short due to variable transaction sizes, network congestion, and the practical constraints of batch preparation and submission.
How do bridge costs compare to internal transaction costs?
Bridge operations between L1 and L2 are generally more expensive than internal L2 transactions because they require execution on both layers and involve additional verification steps. Costs vary significantly depending on the specific bridge implementation and current network conditions.
Conclusion: The State of Layer 2 Scaling
Our comprehensive analysis reveals that while L2 solutions have made remarkable progress in addressing Ethereum's scalability challenges, there remain significant variations in performance characteristics across different implementations.
Fee structures favor optimistic approaches and Optimium solutions in the short term, while ZK rollups offer stronger security guarantees at higher costs. Throughput limitations appear across all solutions, constrained ultimately by Ethereum's data availability capacity.
The upcoming EIP-4844 upgrade represents a potential watershed moment for L2 scalability, promising to reduce costs dramatically while enabling higher throughput across all rollup types. This development, combined with ongoing improvements in proof generation and data compression, suggests a bright future for Ethereum's scaling ecosystem.
👉 Get advanced scaling methodology
As the landscape continues to evolve, users and developers should consider both current performance characteristics and future roadmaps when selecting L2 solutions for their specific needs. The trade-offs between security, cost, and throughput remain significant factors in these decisions.