Paradigm Shift in Ethereum Execution Layer Scaling: From Defensive Conservatism to Empirical Science-Driven Evolution of the 60M Gas Limit

These efforts have enabled the Ethereum mainnet to move from being cautious about raising the Gas Limit to now being able to safely increase the limit all the way up to 60M Gas, or even higher.

Written by: ZHIXIONG PAN

Over the past year, the Gas Limit of Ethereum blocks has rapidly increased from about 30 million to 60 million. This leap has been driven by multiple factors, including protocol-level controls on the worst-case block size, significant optimizations in execution client performance, and systematic testing and verification for higher Gas Limits.

Simply put, developers have reduced the risks of raising the Gas Limit by improving Ethereum protocol rules, significantly increased the speed at which clients process large blocks, and demonstrated that the network can still produce and propagate blocks on time under higher loads.

These efforts have enabled the Ethereum mainnet to move from being cautious about raising the Gas Limit to now being able to safely increase the limit all the way up to 60M Gas. Below, we will explain the concept and history of the Gas Limit in detail, then delve into the core reasons for its recent increase, and look ahead at the conditions required for further scaling in the future.

Gas Limit and Blob: Definition and Differences

The Gas Limit is a parameter in Ethereum that measures the maximum computational workload per block, i.e., the upper limit of the total Gas that can be used for transaction execution in each block. The higher the Gas Limit, the more transactions a single block can accommodate, and the greater the on-chain throughput. However, the side effect is that a higher Gas Limit increases the burden on network participants: block proposers need to package and broadcast larger blocks within a fixed block time, and all nodes in the network must download and execute larger blocks, leading to increased network bandwidth and hardware pressure on nodes.

Blob, on the other hand, is a different type of block content introduced to expand Ethereum’s data availability. Blob originates from the EIP-4844 proposal and allows a block to temporarily contain a large amount of binary data for Layer 2 use, with its cost measured independently from the Gas consumption of regular transactions. In short, Blob provides additional space specifically for L2 Rollup data, while the Gas Limit measures the upper limit of regular EVM computation. The two are not directly comparable: increasing the number of Blobs mainly affects the capacity for L2 data attached to a block, while increasing the Gas Limit directly increases the computational capacity for L1 transaction execution.

This article focuses on discussing the Gas Limit, and will not elaborate on changes in Blob capacity.

Historical Background: Why Was the Gas Limit Not Raised in the Past?

In the early days, Ethereum was always cautious about raising the block Gas Limit. After EIP-1559 was implemented in 2021, Ethereum set the block Gas target at about 15 million (with a single block maximum of about 30 million), and did not increase it for years. The reason was that several key bottlenecks had not yet been resolved, and rashly raising the Gas Limit could threaten network security and decentralization:

• Execution performance: Can client software execute more transactions fast enough? If blocks are too large and nodes cannot complete execution and validation within the block interval, they may miss timely block production or cause chain forks.
• Network propagation: Larger blocks need to be broadcast to the entire network within the 12-second block interval, especially since most validators must receive them within 4 seconds to submit proof of stake on time. Oversized blocks may cause propagation delays, leading to consensus issues.
• State growth: Higher throughput will accelerate the expansion of Ethereum’s global state (ledger data), increasing the burden of node synchronization and storage, which may weaken the network’s decentralization in the long run.
• Hardware requirements: The above factors combined mean that the hardware requirements for running nodes increase. If ordinary users with home computers cannot keep up, a higher Gas Limit may cause the network to concentrate among a few high-performance nodes, which is detrimental to decentralization.

Due to these concerns, the Gas Limit on the Ethereum mainnet remained stable for a long time and did not easily break through the 30 million level. Especially after the rise of Rollups, a large number of transactions compressed data and published it to L1 via low-cost calldata, causing the average Ethereum block size to gradually approach its limit, with single-block data even reaching several megabytes in extreme cases.

Without other improvements, raising the Gas Limit would only further amplify block size and performance issues. Therefore, the Ethereum community at the time chose to mainly rely on Layer 2 scaling, rather than rashly increasing the Gas Limit on L1.

The Core Reasons for the Rapid Increase in Gas Limit Today

So, why has Ethereum been able to safely double the Gas Limit after entering 2025? The fundamental reason is that several technical improvements have been implemented simultaneously, clearing the way for scaling.

Protocol Upgrades Limit Worst-Case Block Size

Ethereum has introduced new protocol rules to reduce the upper limit of “worst-case” block sizes. A key proposal is EIP-7623, which significantly reduces the amount of cheap data that can be included in a single block in extreme cases by increasing the Gas cost of calldata in transactions.

Before EIP-7623 was implemented, attackers could use ultra-low calldata Gas prices to fill a block with up to several MB of data; after the price increase, the same amount of data would consume much more Gas, effectively lowering the block size limit and alleviating the problem of excessive variance between average and extreme block sizes.

This change ensures that even if the overall Gas Limit is raised, the total byte size of blocks will not expand uncontrollably, thus creating a safety margin for increasing the Gas Limit. In other words, the protocol layer has proactively tightened data overhead, ensuring that “doubling computation does not double block size,” laying the foundation for raising the Gas Limit from 30 million to 60 million.

At the same time, the mainnet began to introduce dedicated Blob data transactions for Rollups in EIP-4844, further reducing Rollups’ reliance on cheap calldata. As Rollup data gradually moves from regular Gas space to Blob space, regular block Gas is more focused on actual contract computation, making the average block “lighter,” which also indirectly creates more favorable conditions for raising the Gas Limit.

Significant Optimization of Client Performance

Each Ethereum execution client team has conducted in-depth performance benchmarking and optimization of their software, greatly improving the speed of processing large blocks. The Gas benchmarking framework led by teams such as Nethermind fills entire blocks with a single type of instruction or precompiled contract to stress test the client’s maximum processing capacity (measured in “millions of Gas per second”).

Through this unified benchmark, developers discovered and fixed some previously hidden execution bottlenecks. For example, testing revealed that certain extreme cases of the “Modular Exponentiation” (ModExp) precompile took much longer than its Gas pricing, becoming a common bottleneck for all mainstream clients.

In response, the community quickly proposed EIP-7883 to reprice the ModExp precompile and coordinated client algorithm optimizations. Meanwhile, other time-consuming cryptographic operations (such as BLS12-381 elliptic curve calculations, BN256, hashing, etc.) have also been optimized or repriced by client teams.

According to statistics, after a cross-client “Berlin Interop” performance sprint in mid-2025, the worst-case block processing speed of all execution clients improved significantly, with most operations reaching about 20 million Gas per second.

In other words, if a client can execute 20 million Gas per second, then in the 4-second PoS block interval, it can theoretically process up to 80M Gas per block. This means that raising the block limit to 60M Gas is still within the safety margin.

These performance improvements have eliminated previous concerns that “execution speed cannot keep up with the Gas Limit,” ensuring that even if a block contains twice as many transactions as before, the client can still complete validation within the specified time and will not miss the consensus deadline due to slow execution.

Comprehensive Testing and Verification of Network Propagation Limits

Before implementing any mainnet Gas Limit increase, developers conducted extensive testing on multiple dedicated networks to ensure that larger blocks could still be propagated in a timely manner and accepted by the vast majority of nodes.

For example, in 2025, Ethereum developers raised the block Gas Limit to 60M on the Sepolia testnet and the newly developed Hoodi testnet, and continuously monitored network performance metrics. The results showed that even with blocks using the maximum 60M Gas, block proposals on these networks could still be packaged on time and quickly propagated through the P2P network: 90% of nodes received the block about 0.7~1.0 seconds after block production, and almost all nodes completed validation and accepted the block as the new chain head within 4 seconds.

In other words, even if block Gas usage doubles, blocks can still be propagated across the network before Ethereum’s 4-second validator submission deadline. In these stress tests, developers monitored key data such as whether the proposing node produced blocks on time and the time distribution for all nodes to accept new blocks, and found no significant anomalies.

Since the state size and node topology of testnets differ from the mainnet, developers remain cautiously optimistic, but the test results prove that 60M Gas blocks are feasible both theoretically and technically. At the same time, to ensure consensus layer security, developers also considered Beacon Chain-level limits (for example, the Beacon Chain network layer currently has a ~10MB single-block Gossip propagation limit). By reducing single-block byte size through EIP-7623 and other measures, and avoiding too many penalty transactions occurring at the same time, the 60M Gas execution load does not hit these limits.

Overall, these tests and adjustments have given the core team a full grasp of the risks of raising the mainnet Gas Limit from 30 million to 60 million, boosting their confidence. After most validators expressed support (about 150,000+ validator nodes voted in favor), Ethereum finally began to raise the mainnet Gas Limit in 2025 and plans to officially set the default value to 60M in subsequent upgrades.

Future Outlook: What Is Needed for Further Increases?

The Ethereum community does not intend to stop at 60M Gas. In subsequent upgrade plans such as Fusaka, developers have outlined a path to continue pushing the block Gas Limit to 100M or even higher. To achieve this goal, several technical challenges still need to be solved or continuously monitored:

• Further optimization of heavy computational operations: As mentioned earlier with the ModExp algorithm, the bottleneck has been basically eliminated through EIP-7883 repricing and client optimization. However, to support 100M-level blocks, further optimizations or dedicated accelerations may be needed for other high-Gas cryptographic operations (such as elliptic curve signature verification, zero-knowledge proof verification, etc.). Fortunately, client teams are already collaborating in these areas, and in 2025 tests, they adjusted the implementation of BN256 elliptic curve-related precompiles so that their performance no longer lags behind. It is foreseeable that as Ethereum introduces more high-performance cryptographic primitives (even considering native support for STARKs, etc.), execution bottlenecks will continue to be broken, clearing the way for raising the Gas Limit.
• Controlling state size and node costs: A higher Gas Limit means the on-chain state may grow faster. If not addressed, the difficulty of full node storage and syncing new nodes will increase significantly in a few years. Ethereum developers are already researching state growth issues, such as proposing state rent or periodic pruning of historical state to avoid endless expansion. However, these longer-term mechanisms are still under discussion. In the short term, as the Gas Limit increases, node operators may need to upgrade hardware (such as faster SSDs and more memory) more frequently to keep up with the growing state and data volume. The community emphasizes that raising the Gas Limit will not come at the expense of decentralization, so every step of scaling will be carefully evaluated for its impact on ordinary nodes until state management solutions mature.
• Consensus layer improvements and network protocol optimization: If larger blocks with 100M Gas or more are to be supported in the future, some consensus and network parameters may need adjustment. For example, the current Beacon Chain block has an overall size limit, including execution load, Blob data, and proof data. Developers may need to increase the message size limit at the P2P layer or use compression, sharded propagation, and other technologies to reduce the latency of large blocks. In addition, Ethereum is introducing PeerDAS (peer-to-peer data sampling network) to handle efficient Blob data propagation, which will partially relieve the pressure on execution layer block propagation. After ensuring the safe operation of the execution layer at 60M+ Gas, improvements to the data and network layers will become the next focus of scaling.

Looking ahead, as long as improvements in the above areas progress in sync, further raising the Gas Limit on the Ethereum mainnet is not an unattainable goal. Developers have already verified the feasibility of increasing from 36M to 45M and 60M on testnets, and the next step toward 100M is also being planned. It is important to emphasize that the Ethereum community maintains a consistently cautious attitude toward scaling: every increase will be “tested first, then mainnet,” and will only be implemented after confirming that it does not endanger network security or decentralization.

Overall, the significant increase in the Gas Limit over the past year is the result of collaborative innovation in multiple areas: reduced risk at the protocol layer, improved client performance, and test data providing confidence. Supported by these efforts, Ethereum has successfully taken an important step in L1 scaling and laid the foundation for further increasing capacity and supporting more applications in the future.

References