Ethereum is undergoing its most structurally significant transformation since the Merge. The Glamsterdam upgrade — now in late-stage devnet testing after the Soldøgn interop concluded May 2 in Svalbard, Norway — will triple the network's gas limit from 60 million to 200 million, enshrine proposer...
"Ethereum whale accumulation at these levels is historically a leading indicator of a bottom formation. The supply absorption we are seeing now is a positive signal." — Ali Martinez, Blockchain Analyst
Ethereum is undergoing its most structurally significant transformation since the Merge. The Glamsterdam upgrade — now in late-stage devnet testing after the Soldøgn interop concluded May 2 in Svalbard, Norway — will triple the network's gas limit from 60 million to 200 million, enshrine proposer-builder separation into the protocol (ePBS via EIP-7732), and enable parallel transaction execution through Block-Level Access Lists (BALs). Mainnet activation was targeted for H1 2026; developers now view Q3 as more realistic.
Simultaneously, institutional capital is rotating toward Ethereum. Jane Street cut Bitcoin ETF exposure by 71% while nearly doubling its Ethereum ETF position, adding $82 million in ETH fund exposure in Q1 2026. Spot Ethereum ETFs posted $356 million in net inflows in April, reversing a six-month negative streak. Whales accumulated 140,000 ETH ($322 million) in a 96-hour window in early May. The validator set has contracted from over 1.2 million to approximately 897,000 as EIP-7251 consolidation takes effect — reducing network overhead while concentrating stake into larger, more efficient units.
ETH trades at approximately $2,173 as of May 17, 2026. DeFi TVL on Ethereum stands at $45.74 billion, representing roughly 68% of global DeFi TVL.
Glamsterdam is not a routine parameter adjustment. It rewires two fundamental layers of Ethereum's architecture: how blocks are built (consensus) and how transactions are executed (execution). The upgrade bundles three major EIPs that collectively target a 3.3x increase in Layer 1 execution capacity.
The three workstreams:
The combined effect: Ethereum L1 throughput targets 10,000 TPS — roughly matching Solana's effective sustained throughput of 3,000–5,000 TPS under real-world load, according to The Defiant.
Today, Ethereum validators outsource block construction to specialized builders through MEV-Boost, a system maintained by Flashbots. This process relies on external relays — third-party intermediaries operating on a "trust me" basis with no cryptographic guarantee. According to the EIP-7732 specification, approximately 90% of Ethereum blocks currently flow through this off-chain pipeline.
ePBS replaces this with an in-protocol commit-reveal flow. The block proposer commits to an execution payload header from a builder, which is then revealed and validated within the consensus mechanism. This achieves two things: it removes the relay as a single point of failure and trust assumption, and it expands the data propagation window from approximately 2 seconds to roughly 9 seconds.
The expanded propagation window is not incidental — it is the prerequisite that allows the gas limit to safely increase to 200 million. Without ePBS, a 200M gas limit would risk missed slots and attestation failures as validators struggle to propagate larger blocks within the current 2-second window.
The gas limit increase from 60 million to 200 million is the most visible Glamsterdam metric, but it is only meaningful in conjunction with BALs (EIP-7928). Block-Level Access Lists require transactions to pre-declare which state they will read and write. Nodes can then identify non-conflicting transactions and distribute them across CPU cores for parallel processing.
According to analysis published by BlockEden.xyz, this parallel execution model enables:
If network demand does not rise proportionately to the capacity increase, Ethereum mainnet gas fees could remain near zero for an extended period, per MEXC Research. This creates a direct competitive challenge to Layer 2 networks, which have been subsidizing transactions at median fees of $0.02–$0.06 per transaction.
Over 100 Ethereum core contributors gathered in Longyearbyen, Svalbard — above the Arctic Circle — for the Soldøgn Interop, concluding May 2, 2026. According to the Ethereum Foundation's official recap, the group achieved its three core objectives:
By the final day, nearly all execution clients were running together on the devnet with the external builder pipeline tested end-to-end, per the Foundation blog. The Lambda Class team separately confirmed that the ethrex execution client successfully participated in cross-client interoperability testing during the event.
The Soldøgn results moved the upgrade from "research prototype" to "engineering integration." Public testnets and dual audit phases are planned for mid-2026, creating the earliest realistic path toward safe mainnet activation in Q3 2026.
Ethereum's validator count has fallen from a peak above 1.2 million in mid-2025 to approximately 897,000 as of May 13, 2026, according to beaconcha.in data. The decline is structural, not a sign of network distress.
EIP-7251, implemented in the Pectra upgrade (May 7, 2025), raised the maximum effective balance per validator from 32 ETH to 2,048 ETH. An operator running 1,000 validators at 32 ETH each can now consolidate into 16 validators at 2,048 ETH each — identical total stake, 98.4% fewer validator instances.
The Ethereum Foundation has stated that networking issues emerge as the validator count approaches 1.1 million, with simulations showing serious problems beyond 1.4 million. The consolidation from 1.2 million to 897,000 represents a controlled reduction that improves network health.
Lido, the largest liquid staking protocol, is estimated to begin supporting larger MAX_EB validators and consolidations fully by mid-2026, according to Lido's published roadmap. Approximately 37 million ETH — roughly 30% of circulating supply — remains staked.
A notable milestone: Ethereum validators finalized block 25,000,000 on May 1, 2026, nearly 11 years after the network's genesis on July 30, 2015.
Q1 2026 13F filings reveal a measurable rotation of institutional capital from Bitcoin to Ethereum exposure:
ETH is trading at approximately $2,173 as of May 17, having declined from $2,246 on May 15. DeFi TVL on Ethereum recovered to $45.74 billion in early May, representing approximately 68% of global DeFi TVL.
The institutional thesis centers on the upcoming capacity expansion. A network tripling its throughput while maintaining its dominance in DeFi TVL and stablecoin settlement presents a different risk profile than a static store-of-value asset.
On May 12, 2026, the Ethereum Foundation announced a generational leadership change within its Protocol Cluster. Tim Beiko, Barnabé Monnot, and Alex Stokes — the trio that coordinated Ethereum's protocol development through the Merge, Shapella, Dencun, and Pectra upgrades — are departing. Beiko and Monnot are leaving the organization; Stokes is taking a sabbatical, according to The Block.
The new co-leads:
The Foundation stated that all three new leads contributed to the Fusaka upgrade, PeerDAS data sampling, and the mainnet gas-limit increase. The restructuring reflects Ethereum's increasing specialization: the network's development can no longer be overseen by generalist coordinators; it requires domain expertise in zero-knowledge proofs, post-quantum cryptography, and institutional-grade security.
A tripled L1 gas limit creates a paradox for Ethereum's 73 active Layer 2 rollups, which collectively hold $48 billion in TVL as of May 2026. If L1 fees approach zero, the economic rationale for routing transactions through L2s weakens.
Current L2 median transaction fees, per DefiLlama:
| Network | Median Fee | |---------|-----------| | Base | $0.02 | | OP Mainnet | $0.03 | | Arbitrum One | $0.04 | | zkSync Era | $0.05 | | Scroll | $0.06 |
Arbitrum One leads with $15.9–16.9 billion TVL (40–44% of L2 market), followed by Base at $10.7–12.8 billion. Together they hold approximately 77% of all L2 DeFi liquidity.
ZK rollups are already operating at thin margins. Linea spent $0.012 per transaction in blob fees during Q1 2026; Arbitrum spent $0.018, according to on-chain data. If Glamsterdam pushes L1 fees below these L2 data availability costs, some rollup operators may face negative unit economics on settlement.
The counterargument: L2s provide application-specific customization, higher raw TPS for specialized use cases, and independent governance. But the economic moat narrows substantially with each L1 capacity increase.
While Glamsterdam plays offense on performance, Hegotá — Ethereum's second 2026 upgrade, slated for late 2026 — plays defense. Its scope, defined in the May 2026 Protocol Cluster update:
According to CoinDesk, the FOCIL prototype has a runnable implementation. Hegotá addresses state bloat and geopolitical censorship resistance — structural requirements for a network that aspires to settle institutional-scale value.
Ethereum's position in May 2026 is defined by convergence: a protocol-level capacity tripling arriving alongside measurable institutional capital rotation and a deliberate organizational restructuring. The data points are individually significant; collectively, they represent the network's transition from a constrained, expensive L1 to a high-throughput settlement layer competing directly with both its own L2 ecosystem and rival L1 chains.
The economic implications are unresolved. A 200M gas limit may strengthen Ethereum's value capture by making L1 competitive again for direct transaction settlement — or it may compress L2 margins without proportionally increasing L1 fee revenue. The validator consolidation improves network efficiency but concentrates operational control among fewer, larger operators. The institutional rotation from BTC to ETH ETFs reflects a thesis on Ethereum's utility value rather than its monetary premium — a bet that throughput expansion translates to increased economic activity on the network.
The Soldøgn devnet results — nearly all clients running together, gas limit consensus reached, ePBS tested end-to-end — suggest the engineering is ahead of the market's pricing. Whether Q3 mainnet activation materializes depends on audit outcomes and public testnet stability. The data supports cautious optimism on execution; the market has yet to fully price the structural implications.