Ethereum's scalability, often measured in Transactions Per Second (TPS), is a topic of significant interest. This article analyzes the five primary aspects influencing Ethereum's TPS, providing a clear understanding from its gas mechanism to future upgrades like sharding.
Understanding the Core Constraint: Gas
Unlike Bitcoin, where block space is the primary limited resource, Ethereum's most precious resource is computational gas. The Ethereum Virtual Machine (EVM) allows for complex, programmable transactions. To prevent malicious actors from deploying infinite loops that could crash the network, every operation on Ethereum consumes gas.
Each block has a gas limit, capping the total computational work it can contain. Since block times are relatively fixed, this gas limit directly constrains the network's TPS. When demand for complex transactions is high, the network becomes congested.
The key factors impacting Ethereum's TPS are:
- The Gas mechanism
- Transaction cost structure
- Network infrastructure
- The Merge to Proof-of-Stake
- The implementation of sharding
1. The Ethereum Gas Mechanism
Gas is the unit that measures the computational effort required to execute operations on the Ethereum network.
The Role of Gas and Wei
To incentivize miners (and now validators) to process transactions, users must pay fees. Ethereum defines its currency as ETH, with Wei being the smallest denomination (1 ETH = 1e18 Wei). The name "Wei" honors Wei Dai, a cryptographer known for his contributions to cryptocurrency concepts.
Gas fees are typically denoted in Gwei, which is equal to 1e9 Wei. A transaction's total cost is calculated as:Gas Units Used * Gas Price per Unit (in Gwei)
- Gas Price: The amount of Gwei a user is willing to pay per unit of gas. This is like the price per liter of gasoline.
- Gas Limit: The maximum amount of gas a user allocates for a transaction. This is like the size of a car's fuel tank.
Dynamic Block Gas Limit
The total gas limit for a block is dynamic. Post-London upgrade, Ethereum introduced a variable block size. The target size is 15 million gas, but blocks can expand up to 30 million gas based on network demand. This mechanism allows for a degree of automatic scaling during high traffic.
Impact on TPS
The block gas limit is a direct determinant of TPS. A higher gas limit allows more transactions per block, potentially increasing TPS. The dynamic nature of this limit helps the network adapt to fluctuating demand. For real-time network statistics and current gas limits, you can explore advanced on-chain metrics here.
2. Ethereum Transaction Costs
Transaction costs are split into two components: intrinsic and execution gas.
Intrinsic Gas Cost
This is a base cost determined by the transaction's payload, paid regardless of the computation performed. The formula for intrinsic cost is derived from Ethereum's Yellow Paper:
Intrinsic Cost = (Gtxdatazero Nzeros) + (Gtxdatanonzero Nnonzeros) + Gtransaction + Gtxcreate + Gasscesslist
Where (values in Wei):
Gtransaction= 21,000 (base fee for a simple transfer)Gtxcreate= 32,000 (fee for creating a smart contract)Gtxdatazero= 4 (fee per zero byte in data)Gtxdatanonzero= 16 (fee per non-zero byte in data)Gasscesslistaddress= 2,400Gasscessliststorage= 1,900
A standard ETH transfer has the lowest cost at 21,000 gas. Interacting with smart contracts requires more complex computations, resulting in higher gas limits and costs.
Execution Gas Cost
This cost depends on the specific EVM opcodes used during the transaction's execution (e.g., ADD, MUL, SSTORE). Each opcode has a predefined gas cost. The execution continues until either it completes successfully or the allocated gas (Gas Limit - Intrinsic Cost) is exhausted.
Current TPS Reality
With a common block gas limit of 30 million and a block time of ~13 seconds, the theoretical maximum TPS for simple transfers (21,000 gas each) is around 110. However, in practice, most transactions involve smart contracts, which consume more gas. This results in a real-world average TPS often in the teens.
Key EIPs Affecting Gas and TPS
Numerous Ethereum Improvement Proposals (EIPs) have optimized gas costs. Two of the most significant for future scalability are:
EIP-1559: Fee Market Reform
Implemented in August 2021, this proposal introduced a base fee that is burned and a priority fee (tip) for validators. It made transaction fees more predictable and improved the user experience. While it doesn't directly increase TPS, it creates a more efficient fee market, especially beneficial for Layer 2 solutions. It primarily benefits user experience and ETH's economic model rather than raw TPS.
EIP-4488: Reducing Calldata Gas Cost
This draft proposal aims to drastically reduce the gas cost for transaction calldata (a key cost for Layer 2 rollups) and introduce a total calldata limit per block. The goal is to lower transaction costs on Layer 2 solutions by a factor of 5 or more in the short term, accelerating the transition to a rollup-centric scaling model. While crucial for overall ecosystem scalability, its direct impact on Layer 1 TPS is minimal.
Impact of Transaction Costs on TPS: EIPs like EIP-4488 are designed to supercharge Layer 2 TPS by reducing their costs, indirectly relieving mainnet pressure. The focus for Layer 1 TPS increases lies elsewhere, primarily with sharding.
3. Ethereum Network Infrastructure
Ethereum's peer-to-peer (P2P) network is responsible for propagating transactions and blocks between nodes. The efficiency of this network, governed by the devp2p protocol suite, influences how quickly blocks are synchronized across the globe.
Network-Related EIPs
EIPs like EIP-2464 (transaction announcements) and EIP-2481 (request identifiers) have optimized network communication, improving synchronization speed and efficiency.
EIP-4444: Historical Data Expiry
This draft proposal suggests that clients should stop serving historical data (blocks older than one year) over the P2P network. This would significantly reduce the hardware requirements for running a node (disk space and bandwidth) and simplify client code. While it improves network health and decentralization, its direct impact on TPS is negligible.
Impact of Network on TPS: Network optimizations ensure the chain can handle larger blocks more efficiently, which is a prerequisite for future TPS increases but is not the primary scaling factor itself.
4. The Merge: Transition to Proof-of-Stake
The Merge marked Ethereum's historic transition from Proof-of-Work (PoW) to Proof-of-Stake (PoS). The original execution layer (mainnet) merged with the new PoS consensus layer, the Beacon Chain.
What Changed?
- Consensus Mechanism: Mining (PoW) was replaced with validation (PoS). Validators now secure the network by staking ETH.
- Energy Consumption: PoS reduced Ethereum's energy usage by over 99.9%.
- Block Production: Block times became more consistent and slightly faster, averaging around 12 seconds compared to PoW's ~13 seconds.
Impact on TPS
Crucially, The Merge did not significantly increase Layer 1 TPS. The fundamental block gas limit structure remained the same. The primary goals were improved sustainability, security, and paving the way for future scalability upgrades like sharding. The slight reduction in block time offers a marginal TPS increase, but the order-of-magnitude improvements are dependent on other upgrades.
5. The Future: Sharding
Sharding is the long-term scaling solution for Ethereum's Layer 1. It involves splitting the database horizontally into multiple chains called "shards," distributing the network's load.
Sharding Implementation (Planned)
The current plan focuses on data sharding as the first step.
- Phase 1 (Data Availability): Initially, 64 new shard chains will be created. They will not execute transactions or smart contracts. Instead, their sole purpose is to provide massive amounts of cheap data availability for Layer 2 rollups.
- Synergy with Rollups: Rollups will post their compressed transaction data to these shards. With exponentially more data capacity, rollups can become much cheaper and support vastly higher TPS, potentially reaching 100,000+ TPS for the entire ecosystem.
Impact on TPS
Sharding represents the final piece of the puzzle for Ethereum's scalability roadmap. By providing a scalable data layer for Layer 2 solutions, it enables an enormous effective TPS boost for users. The debate on whether shards will later gain smart contract execution capabilities ("execution sharding") is ongoing, but the data availability focus provides the most immediate and significant scalability gains.
Frequently Asked Questions
What is the main factor limiting Ethereum's TPS?
The primary limit is the block gas limit. Each block can only contain a fixed amount of computational work (gas). Since block times are relatively constant, this caps the number of transactions processed per second.
Did The Merge to Proof-of-Stake increase TPS?
No, The Merge did not significantly increase Layer 1 TPS. Its goals were to change the consensus mechanism to Proof-of-Stake, drastically reduce energy consumption, and set the stage for future scalability upgrades like sharding. Any TPS increase was marginal.
How will sharding improve TPS?
Sharding will improve effective TPS enormously by providing cheap data availability for Layer 2 rollups. Rollups will post their data to shards, allowing them to process transactions off-chain much more cheaply and at a scale aiming for 100,000+ TPS across the ecosystem.
What is the difference between Layer 1 and Layer 2 TPS?
Layer 1 TPS refers to transactions settled directly on the Ethereum mainnet. Layer 2 TPS refers to transactions processed on secondary frameworks (like Optimism or Arbitrum) that are ultimately batched and settled on Layer 1. The future of Ethereum scaling relies on high Layer 2 TPS powered by a secure Layer 1.
What is EIP-1559 and did it help TPS?
EIP-1559 reformed the fee market by introducing a burned base fee and a tip for validators. It made gas fees more predictable and improved the user experience but did not directly increase the network's transaction capacity (TPS). It was a UX upgrade, not a scaling upgrade.
When will sharding be implemented?
The implementation of data sharding is expected to follow The Merge. While timelines in development are always subject to change, it is a primary focus for Ethereum core developers. For the latest updates on protocol changes and their expected impact, you can get advanced methods for tracking development here.