The upcoming year will be pivotal for Ethereum’s scaling efforts. In 2026, the Glamsterdam fork will introduce advanced parallel processing capabilities and elevate the gas limit to 200 million, a significant increase from the current 60 million.
A notable number of validators are expected to transition from reexecuting transactions to verifying zero-knowledge (ZK) proofs. This shift positions Ethereum layer 1 to potentially scale to 10,000 transactions per second (TPS) and beyond, although this goal may not be achieved within 2026.
Concurrently, data blobs may increase (potentially exceeding 72 per block), allowing layer 2s (L2s) to handle hundreds of thousands of transactions per second. L2s are also becoming more user-friendly; for instance, ZKsync’s recent Atlas upgrade permits funds to remain on the mainnet while trading within the swift execution environment of ZKsync’s Elastic Network.
The proposed Ethereum Interoperability Layer will facilitate seamless cross-chain functionality among L2s, prioritizing privacy while enhancing censorship resistance as part of the Heze-Bogota fork at year’s end.
Ethereum in 2026: The Glamsterdam Fork
Ethereum developers are currently finalizing which Ethereum Improvement Proposals (EIPs) will be part of the Glamsterdam hard fork, anticipated in mid-2026. Confirmed changes include Block Access Lists and Enshrined Proposer Builder Separation. Although these titles might not excite, they have the potential to significantly enhance blockchain performance ahead of the transition to ZK technology.
Eventually, core developers may devise more captivating names for features, but for now, we’re left with somewhat mundane technical terminology.
Glamsterdam: Block Access Lists (EIP-7928)
Despite the term “block access lists” suggesting censorship, the upgrade actually facilitates “perfect” parallel block processing.
Historically, Ethereum has functioned in a single-file manner, executing transactions sequentially. Block Access Lists will transform this setup into a multi-lane highway, allowing multiple transactions to be processed simultaneously.
The term refers to a map created by the block producer that details which transactions influence others, accounts, and storage slots, along with state changes post-transaction. This enables transactions to be grouped and executed simultaneously on multiple CPU cores without conflicts.
Related: Blockchains quietly prepare for quantum threat as Bitcoin debates timeline
“With Block Access List, we acquire all the state changes from transaction to transaction and embed that information within the block,” noted Gabriel Trintinalia, a senior blockchain engineer at Consensys working on the execution client Besu.
This method also allows clients to preload necessary data from disk into memory upfront, rather than sequentially reading the disk, addressing what Trintinalia identifies as “the biggest bottleneck we have.”
Effective parallel processing will permit Ethereum to achieve higher transaction rates and accommodate larger block sizes without increasing the gas limit.
Glamsterdam: Enshrined Proposer Builder Separation
The separation of block builders and proposers has commenced with MEV Boost, an external solution employing centralized relays as intermediaries and facilitating about 90% of blocks. Enshrined Proposer Builder Separation (ePBS) integrates this mechanism into Ethereum’s consensus layer for trustless operation.
The rationale behind this separation is for block builders to compete in selecting and ordering transactions optimally while proposers decide which block to present. The goal is to alleviate the centralizing pressure of maximal extractable value (MEV) while enhancing security, decentralization, and censorship resistance.
From a scalability perspective, the significant advantage of ePBS is that it affords more time for the generation and dissemination of ZK-proofs across the network. Currently, validators incur penalties for delays, discouraging them from waiting to validate ZK-proofs. ePBS will allow more time for receiving and validating these proofs.
This can provide attesters more time for receipt and provers additional time for generation, explained Ethereum researcher Ladislaus von Daniels, noting that ePBS decouples block validation from block execution, thus providing another variant of delayed execution.
“This makes opt-in zkAttesting much more incentive compatible for validators.”
Ethereum Foundation researcher Justin Drake estimates that approximately 10% of validators will switch to ZK post-implementation, enabling further increases to the gas limit.
Ethereum L1 Gas Increase and L2 Blob Target Upgrades
The gas limit, currently set at 60 million, is expected to rise significantly in 2026, although predictions vary on the potential new heights.
“In 2026, I would anticipate reaching 100 million fairly quickly. Anything beyond that is probably much too uncertain to predict,” asserted Gary Schulte, senior staff blockchain protocol engineer on the Besu client. He added that the transition to delayed execution could facilitate higher gas limits.
Related: Ethereum tripling its gas limit is the ‘floor, we can go higher’ — Sassano
Tomasz Stańczak, co-director of the Ethereum Foundation, mentioned at the recent Bankless Summit that the gas limit is projected to reach 100 million in the first half of 2026, with an expectation that it could double to 200 million following ePBS. Further enhancements may allow for upwards of 300 million gas per block by year’s end.
Ethereum creator Vitalik Buterin mentioned a more cautious outlook in late November, stating to “expect continued growth but more targeted / less uniform growth for next year. For example, one possible scenario is a 5x gas limit increase simultaneous with a 5x gas cost increase for operations that are relatively inefficient to process.” Buterin pointed to factors such as storage, precompiles, and contracts with large sizes.
Ethereum 2026 Fork No. 2: Heze-Bogota
Some EIPs deferred from Glamsterdam are expected to appear in this fork, but as per Forkast, currently, the only EIP on the Considered for Inclusion list is Fork-Choice Inclusion Lists (FOCIL). Initially proposed for Glamsterdam, it was postponed after intense discussions, as it would have necessitated excessive effort and complexity.
This primarily focuses on enhancing censorship resistance in alignment with cypherpunk principles by enabling multiple validators to dictate the inclusion of certain transactions within each block.
“That is a censorship resistance mechanism that ensures that if at least you have part of the network that’s honest … then you’re going to have your transaction included at some point,” Trintinalia explained.
Stay tuned for part 2, where we will explore scaling L1 using ZK-proofs in 2026.
Magazine: Big questions: Would Bitcoin survive a 10-year power outage?
