Vitalik Buterin announces Ethereum expansion roadmap with Glamsterdam upgrade and ZK-EVM

Vitalik Buterin has announced Ethereum’s new scalability roadmap, divided into two clear phases: short-term and long-term. The plan focuses on optimizing transaction processing, controlling state growth, and gradually transitioning to a proof-based validation model using ZK proofs.

Short-term upgrades in Glamsterdam

In the immediate future, improvements will be implemented in the Glamsterdam upgrade, with three main pillars:

First, block-level access lists enable blocks to be verified in parallel, significantly reducing processing latency.

Second, ePBS (Enshrined Proposer-Builder Separation) not only restructures the block construction mechanism but also allows most of each slot’s time to be used for block verification, instead of just a few hundred milliseconds as currently.

Third, the re-pricing gas mechanism adjusts the cost of operations to more accurately reflect actual resource consumption. Alongside this, Ethereum is beginning to deploy multidimensional gas—an advanced gas model that limits each resource type separately, preventing over-exploitation of any single resource and reducing systemic risk.

Separating “state creation gas”

A key step in Glamsterdam is separating the cost of creating new state from execution and calldata costs.

Currently, an SSTORE operation changing a value from 0 to non-zero costs 20,000 gas, while changing from non-zero to another non-zero costs only 5,000 gas. After the upgrade, this “overhead” could increase significantly (e.g., up to 60,000), but will be accounted for in a separate type of gas called state creation gas.

Notably, state creation gas will not count toward the 16 million gas limit per block. This allows deploying larger smart contracts without inflating the overall network state size.

To ensure compatibility with the EVM, which supports only one-dimensional gas, the development team will implement an N+1 dimensional gas mechanism, including a “reservoir” buffer zone. During execution, the system prioritizes consuming dedicated gas; if insufficient, it draws from the reservoir. This mechanism guarantees two principles: the gas passed into sub-calls remains usable for all purposes, and the GAS opcode always accurately reflects remaining gas after a function call.

In the next phase, Ethereum will adopt a multidimensional pricing model, allowing each resource type to have independently fluctuating gas prices. According to Buterin, this forms the foundation for long-term economic sustainability of the network.

Long-term expansion: ZK-EVM and blobs

For the long term, Ethereum’s scaling strategy revolves around two components: ZK-EVM and data blobs.

Increasing blob throughput with PeerDAS

Ethereum will continue optimizing PeerDAS, aiming for a capacity of around 8 MB of data per second. The goal is not to become a global data layer but to meet the specific needs of the Ethereum ecosystem.

Currently, blobs mainly serve Layer 2 solutions. In the future, Ethereum block data could be directly stored in blobs. Combined with ZK-SNARK proofs, this allows validators to verify data validity and availability without downloading and re-executing the entire chain—an essential step toward supporting a “hyperscaled” Ethereum.

Deployment roadmap for ZK-EVM

Buterin proposes a phased approach to increase confidence in ZK-EVM:

By 2026, client implementations enabling network participation as attesters based on ZK-EVM will emerge. However, the system will not yet be fully secure for the entire network to rely on. An acceptance threshold of around 5% of the network operating in this mode is considered acceptable.

By 2027, Ethereum will recommend a higher percentage (e.g., 20%) of validators using ZK-EVM, along with increased formal verification efforts to strengthen security. Even if only 20% of the network adopts ZK-EVM, the overall gas limit can increase significantly, as solo stakers—who make up less than 20%—will have a lower-cost validation pathway.

Once the system matures, Ethereum will transition to a “3-of-5 mandatory proving” model, where a block is valid only if it contains at least 3 out of 5 different proof types from various verification systems. At this stage, most nodes (except index nodes) will rely on ZK-EVM proofs rather than re-executing the entire block.

Simultaneously, ZK-EVM will continue to be optimized and formally verified at the highest levels. More profound changes to the virtual machine, including potential architecture shifts like RISC-V, may also be considered in the future.

Moving toward a sustainable hyperscale Ethereum

Overall, Buterin’s roadmap aims not only to increase throughput but also to balance execution expansion with control over state growth—key factors for maintaining decentralization.

This strategy indicates Ethereum is gradually progressing toward a “hyperscale” model based on cryptographic proofs and resource separation, rather than simply raising gas limits like traditional chains.