Ethereum's market dominance has fallen to 10.4%, a three-year low, while its flagship Glamsterdam hard fork — targeting a 78% gas fee reduction and 10,000 TPS throughput — remains stuck in devnet stabilization with no confirmed mainnet date. The upgrade's two headline proposals, EIP-7732 (enshrin...
"2026 is the year that we take back lost ground in terms of self-sovereignty and trustlessness." — Vitalik Buterin, Co-Founder, Ethereum
Ethereum's market dominance has fallen to 10.4%, a three-year low, while its flagship Glamsterdam hard fork — targeting a 78% gas fee reduction and 10,000 TPS throughput — remains stuck in devnet stabilization with no confirmed mainnet date. The upgrade's two headline proposals, EIP-7732 (enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists), represent the most structurally ambitious changes to the network since the Merge. But implementation complexity, scope creep from 25+ candidate EIPs, and public warnings from the Base engineering team have pushed realistic deployment expectations from Q2 into Q3 or later.
The stakes are quantifiable. Solana surpassed Ethereum in weekly dApp revenue for five consecutive weeks in April 2026, generating $16.94 million versus Ethereum's $13.55 million, according to DefiLlama data. Polymarket assigns a 60% probability to ETH losing its #2 market cap position. Ethereum's $55 billion DeFi TVL still dwarfs competitors, but the network's L1 fee revenue — the metric that drives ETH burn and staker yield — continues to erode. Glamsterdam is the protocol-level response. Whether it ships in time to reverse the slide is an open question.
Ethereum's share of total crypto market capitalization dropped to 10.4% in April 2026, down from approximately 18% at the start of 2025, according to Phemex data. ETH trades near $2,330 with a market cap of roughly $281 billion. Q1 2026 saw a 27% price decline — steeper than Bitcoin's 20% drawdown over the same period.
Four structural forces drive the compression: Layer 2 networks cannibalizing mainnet fee revenue; competing L1s capturing developer and user activity; institutional capital flowing to Bitcoin ETFs rather than ETH products; and a prolonged period without meaningful L1 performance upgrades. The last major fork, Pectra, shipped in May 2025. Twelve months without a mainnet upgrade is a long gap for a network under competitive pressure.
Glamsterdam bundles two headline EIPs targeting the consensus and execution layers simultaneously. The combined effect, if achieved, would represent the largest single-fork performance gain in Ethereum's history:
Target metrics: gas limit expansion from 60 million to 200 million per block, throughput of approximately 10,000 TPS (up from double-digit TPS currently), and a projected 78.6% reduction in gas costs for smart contract calls.
Today, Ethereum block production operates through an informal but deeply entrenched arrangement. Validators (proposers) outsource block construction to specialized builders via MEV-Boost relay infrastructure. According to Flashbots data, two builders produce roughly 90% of all Ethereum blocks — a centralization level that directly undermines the network's credibility as neutral infrastructure.
EIP-7732 "enshrines" this separation into the protocol itself. Proposers and builders become distinct, protocol-defined roles acting in sequence within consensus. The change has cascading implications: the protocol must now handle disagreement, failure, or censorship between two parties per block slot. As the Ethereum Foundation's Protocol Support Team noted in its April 10, 2026 Checkpoint #9, the implementation requires "every part of the stack to reason about partial blocks and two-party coordination, a change that touches practically everything."
The payoff is structural: by moving builder selection on-chain, the protocol can enforce censorship resistance guarantees, distribute MEV more equitably, and reduce reliance on trusted relays. Flashbots has already begun transitioning to BuilderNet, a TEE-based decentralized building network jointly operated with Beaverbuild and Nethermind. ePBS would formalize and extend this transition at the protocol level.
EIP-7928 addresses a different bottleneck. Today, Ethereum processes transactions sequentially — each transaction executes after the previous one completes, because the EVM cannot predict which state slots a transaction will touch until runtime. This serialization caps throughput regardless of hardware improvements.
Block-Level Access Lists (BALs) require block builders to declare, at the block header level, exactly which accounts and storage slots the block's transactions will access. This allows nodes to pre-fetch data and execute non-conflicting transactions in parallel. The mechanism is analogous to how modern databases use row-level locking to permit concurrent operations.
Together with the gas limit increase from 60 million toward 200 million, BALs create the conditions for a throughput jump from current levels (roughly 15-30 TPS) toward the 10,000 TPS target. According to Tomasz Stanczak, co-director of the Ethereum Foundation, the gas limit would increase to 100 million in H1 2026 and is projected to double to 200 million following ePBS deployment.
The gas limit roadmap is a two-phase process:
| Phase | Target Gas Limit | Expected Timeline | Prerequisite | |-------|-----------------|-------------------|--------------| | Phase 1 | 100 million | H1 2026 | Validator coordination | | Phase 2 | 200 million | Post-ePBS | EIP-7732 deployment |
The current 60 million limit was already a community-driven increase from 36 million in early 2025. Each expansion requires validator coordination rather than a hard fork, but ePBS fundamentally changes the safety assumptions for higher gas limits by restructuring how block production handles resource allocation.
A parallel effort, EIP-7904, introduces benchmarked gas repricing to align opcode costs with actual computational resources. Several opcodes are currently mispriced relative to their real CPU, memory, and I/O costs — a legacy of the original EVM design. Correcting these prices is necessary to prevent the expanded gas limit from creating new attack vectors.
As of the Ethereum Foundation's Checkpoint #9 (April 10, 2026):
The Foundation's stated timeline targets H1 2026. Multiple independent analyses suggest Q3 is more realistic. No mainnet fork date has been announced. The process requires stable devnet, client releases, security reviews, public testnet deployment on Holesky and Sepolia, testnet confirmation, and then mainnet date announcement.
Glamsterdam exists in the context of a structural revenue challenge. Since EIP-4844 (March 2024) introduced blob space for rollup data, L2 settlement costs collapsed. Blob base fees spent most of 2024 and 2025 near 1 wei — effectively zero. Rollups obtained near-free data availability, and Ethereum's L1 captured almost none of the value from L2 activity growth.
Current fee dynamics:
The Glamsterdam thesis is that a faster, cheaper L1 creates a feedback loop: reduced L1 costs attract direct activity back to mainnet (higher-security settlement); simultaneously, increased L2 usage generates more blob demand, eventually pushing blob fees above the 1-wei floor. Industry estimates suggest blob fees could contribute 30-50% of total ETH burn by late 2026, contingent on L2 activity scaling and blob capacity reaching 48 blobs per block by mid-2026.
The urgency is underscored by competitive data. Solana processed 25.3 billion transactions in Q1 2026, per CoinTelegraph, versus Ethereum's 200 million. Solana's average transaction cost: $0.00025. Ethereum mainnet: $0.50-$3.00.
In February 2026, Solana generated $26.7 million in monthly network revenue versus Ethereum's $23.2 million. For the week ending April 21, 2026, Solana dApp revenue reached $16.94 million compared to Ethereum's $13.55 million — the fifth consecutive week Solana led, according to DefiLlama.
Ethereum's structural advantage remains in capital depth: $55 billion in DeFi TVL (55-60% of global DeFi) versus Solana's approximately $8 billion. Aave V3 alone holds $26.18 billion; Lido holds $23.07 billion. But TVL is a stock metric. Revenue is a flow metric. The flow is moving toward competitors.
Scope creep. Over 25 non-headliner EIPs are under consideration for inclusion. The Base engineering team has publicly warned that adding FOCIL (Fork-Choice Inclusion Lists) alongside ePBS could push the upgrade beyond 2026. FOCIL was subsequently moved to Hegota, the planned H2 2026 fork, but scope discipline must hold through testnet.
ePBS complexity. The protocol change requires every consensus and execution client to handle partial blocks and two-party coordination. This is not incremental complexity; it is architectural. Client team velocity on ePBS has been slower than anticipated.
Hegota dependency. The second 2026 fork (Hegota) carries Verkle Trees, which cut node storage by approximately 90% and enable stateless clients. Any Glamsterdam delay compresses the Hegota window, potentially pushing Verkle Trees into 2027. The Ethereum Foundation aims for a biannual cadence; a slip here disrupts the entire pipeline.
Market window. Each month of delay extends the period where Ethereum lacks the performance improvements that could reverse fee revenue erosion and dominance decline. The competitive landscape does not pause for devnet stabilization.
Glamsterdam is not a performance tweak. It is an architectural restructuring of how Ethereum builds blocks and executes transactions. The 78% gas reduction and 10,000 TPS target, if achieved, would represent the largest single-fork performance improvement in the network's history. ePBS directly addresses a centralization problem — 90% of blocks built by two entities — that undermines the network's core value proposition.
The question is timing. Every month without Glamsterdam is a month where Solana's revenue lead widens, where L2s capture value that L1 does not, and where Ethereum's 10.4% dominance can compress further. The Ethereum Foundation has demonstrated delivery capability — Pectra shipped on schedule in May 2025. But Glamsterdam's scope is categorically different, and the ePBS integration challenge is unlike anything the client teams have previously shipped.
The data does not support a confident prediction on delivery date. What the data does support: Ethereum's competitive position deteriorates the longer this takes.