Ethereum, as one of the most prominent blockchain platforms, powers a vast ecosystem of decentralized applications (dApps), smart contracts, and digital assets. However, its transaction processing capacity—commonly measured in Transactions Per Second (TPS)—remains a critical bottleneck. Despite ongoing upgrades and innovations, Ethereum’s current TPS hovers around 10–30 under normal conditions, far below the throughput required for mass adoption.
This article explores five core factors that directly impact Ethereum’s TPS: the gas mechanism, transaction costs, network infrastructure, the Merge, and sharding. We’ll examine how each element shapes performance, scalability, and user experience—while integrating essential keywords such as Ethereum TPS, gas limit, EIP-1559, layer 2 rollups, blockchain scalability, consensus layer, and network congestion naturally throughout.
1. The Gas Mechanism: Ethereum’s Computational Fuel
At the heart of Ethereum’s architecture lies the gas system—a unique mechanism designed to prevent abuse and ensure fair resource allocation across the network.
Unlike Bitcoin, where transactions are simple value transfers, Ethereum supports complex operations via smart contracts. Each computation—be it a conditional statement, loop, or data write—consumes computational resources. To prevent infinite loops or spam attacks, Ethereum assigns a cost to every operation in units called gas.
Think of gas as fuel for computation:
- A basic ETH transfer costs 21,000 gas
- Smart contract interactions can require hundreds of thousands of gas
- Each block has a gas limit, currently capped at approximately 30 million gas
👉 Discover how blockchain networks manage transaction loads efficiently.
The average block time is about 12 seconds post-Merge, meaning Ethereum can process roughly:
30,000,000 / 21,000 ≈ 1,428 transactions per block
→ ~119 TPS (theoretical maximum)However, real-world TPS is much lower due to mixed transaction types and network dynamics.
Dynamic Block Size & Network Demand
Post-London Upgrade (EIP-1559), Ethereum introduced variable block sizes with a target of 15 million gas per block. Blocks can expand up to double that size during high demand, allowing temporary scalability boosts. This flexibility helps absorb traffic spikes but doesn’t solve long-term congestion.
Crucially, gas is the scarcest resource on Ethereum, not storage space like in Bitcoin. This makes gas efficiency central to improving TPS and reducing user costs.
2. Transaction Costs and Their Impact on Throughput
Every Ethereum transaction incurs two types of gas costs: intrinsic cost and execution cost.
Intrinsic Cost: The Entry Fee
This is the base cost determined by the transaction’s payload:
- Standard ETH transfer: 21,000 gas
- Adding zero-byte data: +4 gas per byte
- Non-zero bytes: +16 gas per byte (post-Istanbul upgrade)
- Contract creation: additional 32,000 gas
If a transaction’s intrinsic cost exceeds its specified gas limit, it’s rejected immediately—no execution occurs.
Execution Cost: The Runtime Expense
Once a transaction begins executing, every EVM opcode consumes predefined gas:
ADD,SUB: minimal costSSTORE(writing to storage): high costCALL: variable based on context
Unused gas is refunded to the sender; consumed gas goes to miners (now validators). The refund system prevents infinite loops by ensuring code execution halts when gas runs out.
EIP-1559: A Game-Changer for Fee Market Efficiency
Introduced in August 2021, EIP-1559 overhauled Ethereum’s fee model:
- Introduced a base fee, dynamically adjusted per block and burned
- Users add a priority fee (tip) to incentivize inclusion
- Eliminated first-price auction inefficiencies
While EIP-1559 improved predictability and reduced wallet complexity, it did not increase TPS. Instead, it optimized fee markets, enhanced UX, and supported Layer 2 scaling by stabilizing cost expectations.
EIP-4488: Reducing Calldata Cost for Rollups
Proposed by Vitalik Buterin, EIP-4488 aims to slash calldata costs from 16 gas per non-zero byte to just 3—while imposing a daily cap to avoid P2P network strain.
This change would significantly reduce fees for layer 2 rollups like Optimism, Arbitrum, and zkSync—potentially lowering their costs by 75–80%. Although beneficial for ecosystem growth, this may slightly decrease L1 TPS due to larger block sizes.
3. Network Infrastructure and Data Propagation
Ethereum relies on a peer-to-peer (P2P) network protocol suite known as devp2p for node communication. Efficient data propagation between nodes directly affects block propagation speed and overall throughput.
Delays in syncing blocks or propagating transactions increase latency and raise the risk of uncle blocks—especially problematic under high load.
EIP-4444: Pruning Historical Data
To reduce node burden, EIP-4444 proposes removing historical data older than one year from execution clients:
- Lowers disk requirements from over 1TB to manageable levels
- Reduces bandwidth usage
- Enables lightweight clients without sacrificing security
While this improves decentralization and node accessibility, it has minimal direct impact on TPS.
P2P Optimization Efforts
Other proposals like EIP-706 (Whisper), EIP-778 (ENR), and EIP-2364 (ETH/65) aim to enhance discovery, encryption, and transaction announcement protocols. These improvements support network resilience but don’t drastically alter transaction throughput.
4. The Merge: Transition to Proof-of-Stake
The Merge marked Ethereum’s shift from Proof-of-Work (PoW) to Proof-of-Stake (PoS), completed in September 2022.
Under PoW, average block times fluctuated around 13 seconds. Post-Merge, with the Beacon Chain serving as the new consensus layer, block intervals stabilized to exactly 12 seconds per slot.
Does the Merge Boost TPS?
Not significantly. While block production became more consistent and energy-efficient:
- Block size remains constrained by gas limit
- No increase in data capacity or execution speed
- TPS saw only marginal improvement
However, the Merge laid the foundation for future upgrades like sharding and Verkle trees by enabling better coordination among validators.
👉 Learn how next-gen blockchain consensus models are reshaping scalability.
5. Sharding: The Long-Term Scalability Solution
Sharding is Ethereum’s ultimate plan to achieve massive scalability by splitting the network into multiple parallel chains—64 shard chains—each handling separate data streams.
Phase 1: Data Availability Shards
Initially, shards will serve as data layers, not execution environments. They won’t process smart contracts but will provide extra data space for rollups.
When combined with Layer 2 rollups:
- Rollups bundle off-chain transactions
- Submit compressed proofs and data to L1
- Shards supply scalable data availability
This synergy could enable Ethereum to reach over 100,000 TPS in the long term.
Future Possibilities: Execution Sharding?
There's ongoing debate about whether shards should eventually run code:
- Keep shards as pure data layers
- Enable partial execution on select shards
- Wait for ZK-SNARKs maturity before deciding
Vitalik Buterin leans toward option 1 for simplicity and security, prioritizing rollup-centric scaling over full execution sharding.
Frequently Asked Questions (FAQ)
Q: What is Ethereum’s current TPS?
A: Under normal conditions, Ethereum processes between 10 and 30 transactions per second, depending on transaction complexity and network congestion.
Q: Why doesn’t increasing the gas limit boost TPS permanently?
A: While raising the gas limit allows more transactions per block, it also increases block propagation time and node hardware requirements—raising centralization risks. Hence, increases are limited and temporary.
Q: How do Layer 2 rollups improve TPS?
A: Rollups execute transactions off-chain and post compressed data on Ethereum. This reduces L1 load while inheriting its security—enabling higher throughput at lower costs.
Q: Will sharding launch in 2025?
A: Sharding deployment depends on post-Merge progress. While full execution sharding may take years, initial data shards could arrive by 2025–2026, accelerating rollup adoption.
Q: Does EIP-1559 increase Ethereum’s TPS?
A: No. EIP-1559 improves fee market efficiency and user experience but does not expand block capacity or execution speed.
Q: Can Ethereum scale without sharding?
A: Partially—via rollups and other Layer 2 solutions—but sharding is essential for sustainable long-term scalability and affordability at global scale.
Final Thoughts: Toward a Scalable Ethereum
Ethereum’s journey toward higher TPS isn’t about one breakthrough—it’s a layered evolution combining protocol upgrades, economic incentives, and architectural innovation.
From the foundational gas mechanism to transformative milestones like the Merge and future-ready designs like sharding, each component plays a role in building a robust, scalable network.
While immediate TPS gains remain modest, the roadmap points toward an ecosystem capable of supporting millions of users through rollup-centric scaling and efficient data distribution.
👉 Explore cutting-edge tools that help developers build scalable dApps on Ethereum today.
As these technologies mature, Ethereum moves closer to fulfilling its vision: a decentralized world computer accessible to all.