In a long-form research report written by ShirleyLi of Web3Caff Research, Ethereum’s future is examined through the lens of nearly twelve years of roadmap changes, governance adjustments and protocol-level redesign. Since Vitalik Buterin and his team formally introduced Ethereum to global users at an international conference in 2014, the network has moved from a niche experiment into one of the most influential base platforms in the Web3 ecosystem. The report frames Ethereum’s current challenge as a constant tension between maintaining stability and continuing to change direction when the ecosystem requires it.
From Frontier to the 2026 Strawmap
Between 2014 and 2016, Ethereum gradually formed an early phased roadmap: Frontier, Homestead, Metropolis and Serenity. The first three stages are usually regarded as Ethereum 1.0, focused on improving basic functions and strengthening network stability. Serenity represented the long-term target: rebuilding the consensus mechanism and core architecture to achieve a leap in scalability and performance. In 2020, Ethereum clarified the Serenity technical path, formally confirming the transition to proof of stake and introducing the logic of sharding, which marked the start of a systematic architectural rebuild.
In 2022, Ethereum released a more complete medium- and long-term roadmap and confirmed a Rollup-centric scaling path. Execution would expand to Layer 2 networks, while the main chain would focus on security and data availability. In February 2026, the Ethereum Foundation again published a draft “Strawmap” for the next decade, setting more specific optimization goals across the consensus layer, data layer and execution layer. The report argues that every roadmap revision reflects Ethereum’s phased balancing act among scalability, security, decentralization and ecosystem value distribution.
Foundation governance and funding changes
At the beginning of 2025, as the wider market environment improved and new narratives emerged, Ethereum’s development pace was viewed by parts of the community as relatively slow. Criticism focused on execution efficiency, market communication and ecosystem expansion by the Ethereum Foundation and core developers. The Foundation then made a series of major adjustments. In February 2025, Aya Miyaguchi, who had served as executive director since 2018, moved into the newly created role of President, with responsibilities shifting toward external cooperation, institutional relations and cultural communication. Nethermind founder Tomasz Stańczak and Hsiao-Wei Wang became co-executive directors.
Under the new structure, the Ethereum Foundation streamlined its organization, cut 19 employees and moved its strategic emphasis from Layer 2 back toward Layer 1 itself. It also placed more emphasis on external communication and transparency around technical direction, development priorities and use of resources. In June 2025, the Foundation reorganized its internal research and development system: “Protocol Research & Development,” or PR&D, was simplified to “Protocol.” The team focused on three short-term goals: scaling L1 performance, scaling Blobs and improving user experience. The adjustment marked a shift from a research-first orientation toward engineering implementation and practical delivery.
In February 2026, Tomasz Stańczak announced that he would step down as co-executive director, with Bastian Aue and Hsiao-Wei Wang taking over. During his term, Stańczak promoted explorations in privacy protection, quantum computing security and the combination of AI and Ethereum. After leaving the role, he would devote more attention to products and infrastructure related to the convergence of AI and blockchain. The latest internal structure lists Vitalik Buterin, Aya Miyaguchi, Patrick Storchenegger and Hsiao-Wei Wang as board members, responsible for governance and strategic direction, while management and functional teams handle execution and operations.
The Foundation also changed its grant system in 2025. In August, it paused the open grant program that had operated since 2018, then restarted a new Ecosystem Support Program, or ESP, in November. After the adjustment, funding moved from passively accepting applications toward active guidance. The first supported areas covered cryptography, privacy, the application layer, security and community growth. The Foundation also decided to reduce annual spending from about 15% to 5% in order to slow the consumption of ETH reserves. In May 2026, Ethereum Foundation researchers Carl Beek and Julian Ma announced departures, while former researcher Dankrad Feist publicly said the ecosystem needed a new organization more aligned with Ethereum’s economic interests to “save” Ethereum. Vitalik Buterin and Ethereum co-founder Joe Lubin both responded that the disputes reflected a phase of adjustment between long-term technical construction and the current commercialization process.
PoS design, staking and Blob capacity
Ethereum’s move to proof of stake ended its high-energy consensus model, but the 32 ETH staking threshold also raised the entry barrier for validators and created a degree of validator centralization risk. Vitalik Buterin previously proposed strengthening security by increasing the participation ratio required for block finality, for example moving from the current roughly two-thirds threshold signature level to 75% or higher. In May 2025, the Pectra upgrade was activated on Ethereum mainnet. EIP-7251 raised the maximum effective validator balance from 32 ETH to 2048 ETH. The 32 ETH minimum requirement to become a validator remained unchanged, but larger stakers no longer needed to split their holdings across many validator nodes to receive corresponding incentives, helping reduce communication and coordination overhead across the network.
Pectra also included EIP-7002, which optimized the staking withdrawal mechanism by introducing execution-layer-triggered withdrawals. Under specific conditions, stakers can complete withdrawals without requiring an active validator signature, improving control over assets and reducing the complexity of joining and exiting staking. The Ethereum Foundation has also explored Distributed Validator Technology, or DVT, which splits a single validator’s private key and signing capability across multiple cooperating nodes to reduce single points of failure. It has tested lighter implementations such as DVT-lite to simplify deployment and operations. According to information disclosed by the Foundation in March 2026, 72,000 ETH had participated in staking through related mechanisms.
EIP-4844, implemented in the Dencun upgrade, introduced a low-cost data availability space called Blobs, mainly for temporary data. Layer 2 networks can submit batched transaction data into Blobs, reducing the cost of publishing data on-chain. The original design set a target of 3 Blobs per block and a maximum of 6. To avoid excessive use, Ethereum introduced a floating fee mechanism: if the number of Blobs in a block exceeds 3, the base fee rises; if it is below 3, the base fee falls to encourage use. As Layer 2 demand for Blob space grew, actual usage increasingly approached or reached the target, triggering fee increases at multiple points.
To address the pressure on Blob capacity, Pectra included two related proposals. EIP-7691 increased the target number of Blobs per block from 3 to 6 and raised the maximum from 6 to 9. EIP-7623 attempted to raise the cost of using Calldata, encouraging Layer 2 networks to use Blobs more and reducing pressure from mainnet data publication. With the Fusaka upgrade in December 2025, Ethereum introduced additional mechanisms at the data availability layer. On December 11, 2025, the Ethereum Foundation stated that Blob capacity per block had been increased to 15.
The report emphasizes that Blob design is not isolated. It is part of Ethereum’s scaling path after the network shifted from execution sharding toward data sharding. Ethereum originally proposed sharding to split the network into multiple subchains that could process transactions and data in parallel, but execution-layer sharding faced high engineering complexity and security challenges. Data sharding became the priority, and Blobs became the transitional implementation. They allow Layer 2 networks to publish batch transaction data to Ethereum mainnet at lower cost, while PeerDAS further optimizes the validation of Blob data through data availability sampling.
Rollups, fragmentation and renewed coordination with Layer 1
As the number of Rollup solutions increases, their relationship with Ethereum mainnet is also changing structurally. Rollups reduce the execution load on mainnet by moving transaction execution to Layer 2, but this also changes the fee distribution structure. More transaction fees paid by users flow to Layer 2, while mainnet mainly handles data publication and settlement. Different Rollups are relatively independent in architecture and ecosystem, creating several parallel sub-ecosystems and increasing the complexity of cross-Rollup interaction. Interoperability protocols such as LayerZero and Hyperlane try to standardize message passing across Rollups, while shared sequencer projects such as Astria and Espresso attempt to offer unified sequencing services for multiple Rollups. The report notes that these “Rollup–third-party infrastructure–Rollup” communication patterns are still non-native Ethereum expansion paths, adding trust assumptions and security issues, while shared sequencers also affect MEV distribution at Layer 1.
Rollups using different technical approaches are building separate ecosystems. Arbitrum is developing Orbit chains through its Nitro / Orbit stack, with Plume Network and ApeChain listed as examples. zkSync supports multi-chain expansion through its Elastic Chain architecture, with Abstract and ZERO Network as representative projects. Optimism has built the Superchain around OP Stack, with members including OP Mainnet, Base, BOB and Soneium. In mid-to-late February 2026, the Base team announced that it would gradually move away from Optimism’s OP Stack architecture and toward its own Base Stack unified technology stack. Because Base was originally built using OP Stack, the report describes this more as customized expansion based on an open-source framework than a complete departure from Optimism. Base also stated that it would continue collaborating with the Optimism ecosystem and follow relevant open-source standards.
The Ethereum Foundation has started addressing this relationship more directly. While continuing to improve Rollup data availability, it established a Platform Team under the Protocol team framework to coordinate between protocol research and ecosystem development. The team is intended to reorganize and optimize the relationship between mainnet and Layer 2, turning the two from a delicate alliance into a more coordinated and mutually reinforcing platform. On March 23, 2026, the Foundation again published a systematic explanation of the evolving roles of Layer 1 and Layer 2. Under the newer definition, Layer 2 is no longer limited to scaling; it also serves differentiated needs that Layer 1 cannot provide, such as stronger privacy, lower latency and compliance adaptation for specific scenarios. Ethereum mainnet is positioned as the core settlement and liquidity layer for a multi-chain ecosystem.
To move toward this goal, Ethereum is pursuing two paths. The first is strengthening the connection between Rollups and mainnet. With Ethereum Foundation co-funding, Gnosis co-founder Friederike Ernst and Zisk founder Jordi Baylina announced the Ethereum Economic Zone, or EEZ. The plan attempts to build an L1<>L2 framework in which cross-chain interactions between Rollups and mainnet, and between Rollups themselves, are brought into a unified execution framework. In that design, cross-chain interaction would not rely entirely on traditional asynchronous message passing. Contract calls across execution environments could be completed within the same execution flow and have atomicity, meaning related operations either all succeed or all roll back, reducing inconsistent state risks.
The second path is Native Rollups, proposed in EIP-8079 through a concept verification prototype. Today, both Optimistic Rollups and ZK Rollups execute transactions, generate state and then use additional mechanisms such as fraud proofs or zero-knowledge proofs to prove correctness to Ethereum mainnet. They must also follow Ethereum’s underlying rules for processing transactions and updating state, which requires maintaining a relatively complex execution and verification system. EIP-8079 attempts to expose Ethereum’s state transition function as an open interface for Rollups. Rollups could submit transactions to mainnet, and mainnet would complete state calculation according to unified rules, reducing reliance on independent proof systems and lowering Rollup maintenance costs.
ePBS, MEV, single-slot finality and FCR
After Ethereum moved to proof of stake, the network randomly selects a block proposer from validators who stake ETH. The proposer is responsible for final block publication and can extract value beyond standard block rewards and gas fees by changing transaction order, a practice known as MEV. To reduce the control that a single validator has over transaction ordering and value extraction, Ethereum introduced the idea of proposer-builder separation, or PBS. Builders collect, order and package transactions into candidate blocks, while proposers select from multiple candidate blocks and submit the final block. This separates value extraction capability from block confirmation authority.
In practice, PBS has not previously been directly written into the protocol. It has been implemented through third-party middleware such as MEV-Boost, allowing validators to outsource block construction to third-party builders and rely on relays for information transmission. While this improved block-building efficiency, it remained an off-chain market, depended on trusted relays, carried centralization risk and lacked on-chain transparency and constraints in MEV distribution. One of the most important planned changes in Glamsterdam is EIP-7732, or Enshrined Proposer-Builder Separation. ePBS writes proposer-builder separation directly into the Ethereum protocol, introduces builders as formal network participants, requires them to stake first and submit block commitments, and decouples consensus validation from execution validation. The network can prioritize consensus-layer validation during critical block propagation and complete full execution-layer validation later.
This adjustment reduces reliance on relays and gives nodes more time to propagate block data, helping Ethereum support larger data loads, especially Blob writes, while maintaining stability. The report also places this alongside competition from high-performance Layer 1 networks. Solana uses parallel execution and local fee markets to improve processing capacity, while Sui uses an object model and DAG architecture to achieve more efficient concurrent processing in certain scenarios. Such designs can serve use cases including high-frequency trading and on-chain games. The report presents this not simply as direct confrontation, but as different infrastructure exploration paths across blockchains.
Ethereum’s long-term answer to confirmation time is single-slot finality, or SSF. In Ethereum, validators reach consensus roughly every 12 seconds on the transactions and order that can be included in a block; this window is called a Slot. Every 32 Slots, about 6.4 minutes, form an Epoch. Under the current mechanism, a block needs at least two-thirds of validators to complete two rounds of voting, or two Epochs of confirmation, to be considered finalized and irreversible. This normally takes about 12 to 15 minutes. SSF aims to compress finality that currently spans two Epochs into a single Slot.
Before SSF is fully implemented, the Ethereum Foundation has proposed a transitional optimization called the Fast Confirmation Rule, or FCR. Its goal is to reduce confirmation times for deposits from Ethereum L1 to L2s and centralized trading platforms from several minutes to about 13 seconds. FCR does not change the existing finality mechanism. Instead, it uses the first round of voting earlier to judge security. If the first vote has already reached a high proportion, such as clearly exceeding two-thirds, the system can treat the block as basically confirmed. The report stresses that FCR is not the same as true finality. Its security depends on most validators behaving honestly and on stable, low-latency network communication. If the network becomes congested or is attacked, a rapidly confirmed block can still be reorganized. The original report also states that its content is for reference only and does not constitute any forecast, investment advice, proposal or offer.