Ethereum entered its 11th year with a sharper division of labor between its base layer and the networks built on top of it: the Fusaka upgrade expanded the chain’s capacity to support rollups, while the Ethereum Foundation cut roughly one-fifth of its staff and reorganized around protocol, user access, and institutional adoption.
Fusaka activated on Ethereum mainnet on Dec. 3, 2025, becoming the network’s second major upgrade of that year after Pectra. Its defining feature, PeerDAS, changes how nodes check whether rollup data is available. Rather than requiring nodes to download every blob — a temporary data format introduced for rollups through 2024’s Dencun upgrade — they can verify sampled portions of that data.
That design gives Ethereum room to carry larger amounts of rollup data without placing the full storage and bandwidth burden on every node. Two parameter-only updates after Fusaka raised the blob target per block from six to 14 and the maximum from nine to 21. Ethereum’s default block gas limit also rose from 45 million to 60 million.
The changes place more of Ethereum’s transaction growth on layer-2 networks while preserving the main chain’s role as the settlement and data-availability layer. Data from growthepie as of July 31, 2026 showed tracked Ethereum L2s processing about 24.61 million transactions daily, compared with 1.77 million transactions on Ethereum mainnet. Weekly and monthly activity showed a similar split, with L2s handling nearly 10 times mainnet transaction volume.
A network designed around rollup capacity
Ethereum’s scaling strategy has evolved through a sequence of upgrades that progressively shifted high-volume activity toward rollups. The network launched as Frontier on July 30, 2015, allowing developers to deploy public smart contracts. Its path since then has included the 2016 DAO hard fork, which created Ethereum Classic; the 2020 launch of the Beacon Chain; and the 2022 Merge, which moved Ethereum to proof-of-stake.
Shapella enabled staking withdrawals in 2023, while Dencun introduced EIP-4844 blob data in 2024. Blobs gave rollups a cheaper way to publish transaction data to Ethereum, reducing a major cost that had constrained layer-2 networks.
Pectra, activated in 2025, brought EIP-7702 account changes and raised the maximum effective validator balance to 2,048 ETH. The larger validator balance limit was intended to reduce operational complexity for major staking operators by allowing them to consolidate validator balances rather than managing a larger number of 32 ETH validators.
Fusaka builds on Dencun’s blob architecture rather than replacing it. PeerDAS is intended to make increased blob capacity practical without forcing every node to retain and verify the complete dataset. That balance will become more relevant if Ethereum continues to raise its data capacity in later upgrades, as its technical plans suggest.
Ethereum’s layer-2 ecosystem has also begun attracting financial-asset projects built around the main chain’s settlement model. Robinhood Chain launched on July 1, 2026, using an Arbitrum-based stack, EVM compatibility and ETH for gas. The network is designed for stock tokens and other financial assets and ultimately settles back to Ethereum.
Mainnet remains the largest pool of on-chain capital
Despite L2s handling the majority of transactions, Ethereum mainnet remained the leading network for stablecoins, decentralized finance and tokenized real-world assets as of late July, according to RWA.xyz and DeFiLlama.
RWA.xyz recorded about $155.9 billion in stablecoins on Ethereum mainnet, out of approximately $296.9 billion tracked globally. That represents about 52.5% of supply. The platform also counted 1,552 tokenized real-world assets on Ethereum, with about $17.15 billion in distributed on-chain asset value across those products.
Ethereum’s RWA total was about 3.3 times that of second-ranked BNB Chain among the 38 networks tracked by RWA.xyz. The figure measures assets directly on Ethereum mainnet and should be separated from ecosystem-wide estimates that include L2s.
Ethereum Institutional, an independent nonprofit that emerged from Ethereum Foundation incubation, said on July 29 that Ethereum and its L2 networks accounted for more than 60% of stablecoin supply and more than 75% of tokenized real-world assets under its broader measurement scope. The organization also announced its first ecosystem funding round and a supporter alliance involving more than 100 ecosystem participants, without disclosing the funding amount.
DeFiLlama placed Ethereum mainnet’s DeFi total value locked at about $41.2 billion, nearly 55% of all TVL across tracked networks. BSC and Tron each held roughly $4.9 billion, leaving Ethereum’s mainnet TVL at around eight times the next-largest chain by that measure.
The figures show the practical result of Ethereum’s scaling approach: lower-cost networks carry much of the transaction count, while mainnet continues to hold a large share of stablecoin liquidity, DeFi collateral and tokenized assets. Higher throughput on L2s does not automatically shift the location of the ecosystem’s largest capital pools.
Foundation reduces staff and redraws its structure
The Ethereum Foundation completed a multi-month restructuring on June 23, 2026, reducing its team by 54 people, or about 20% of its previous headcount. The Foundation has reorganized its work into five areas: protocol, access, user, community and institutional. Those groups operate alongside an operations cluster and management and support teams.
In a mandate published March 13, the Foundation named user autonomy as its central objective. It identified censorship resistance, open source development, privacy and security as its “CROPS” principles.
The organizational overhaul also moved several initiatives outside the Foundation. Ethlabs began operating as an independent nonprofit research-and-development organization. Ethereum Institutional separated from Foundation incubation as a nonprofit, while EthSystems became a for-profit company continuing work associated with an institutional privacy task force.
The Foundation’s treasury policy referenced a 2025 operating-spend level equal to roughly 15% of its treasury. Its stated plan is to reduce that level approximately linearly to 5% over five years. The lower spending target and staff reduction suggest the Foundation is seeking to preserve long-term funding capacity while shifting more specialized work to independent organizations.
Plans extend to faster finality and quantum resistance
Ethereum’s technical agenda has become more explicitly structured following Fusaka. Justin Drake’s “Lean Ethereum” vision, published July 31, 2025, outlined longer-term targets including finality in seconds, about 1 gigagas per second of L1 capacity and 1 teragas per second on L2s. Drake framed those throughput figures as roughly 10,000 transactions per second on the base layer and 10 million transactions per second across L2s.
The plan includes next-generation data sampling, an execution environment designed to work more efficiently with SNARK proofs while retaining EVM compatibility, and a gradual transition toward cryptography resistant to quantum-computing attacks.
EF Architecture’s February 2026 “Strawmap,” developed with input from contributors including Vitalik Buterin, organizes related work around five directions: Fast L1, Gigagas L1, Teragas L2, Post-Quantum L1 and Private L1.
After a Berlin researchers meeting on July 4, Buterin wrote that major portions of the Lean Ethereum agenda could require three to four years and several upgrades. Topics under discussion include recursive STARKs, post-quantum cryptography, redesigned consensus and finality mechanisms, multidimensional gas accounting, new state types and client architecture.
The next planned upgrade, Glamsterdam, combines the execution-layer name Amsterdam with the consensus-layer name Gloas. Developers have targeted the second half of 2026 for planning, though no mainnet activation date had been announced as of July 30.
Glamsterdam’s featured proposals include ePBS, which would embed proposer-builder separation into Ethereum’s protocol, and BAL, a block-level access list that would declare in advance the accounts and storage slots a block expects to access. Testing discussions have linked those changes, along with EIP-8037 repricing, to a potential post-upgrade baseline of a 200 million gas limit.
Beyond Glamsterdam, developers are discussing Hegotá for a possible 2027 timeline. Its selected consensus feature, FOCIL under EIP-7805, would use in-protocol transaction inclusion lists to reduce a single block builder’s ability to exclude transactions repeatedly. Frame Transaction, tracked as EIP-8141, remains under consideration and could separate transaction validation, execution and gas payment into distinct frames, creating a route toward native account abstraction and future post-quantum account migration.
Dive deeper into Ethereum’s evolution and upcoming upgrades in 2025 and beyond with our guide to Ethereum Pectra Upgrade.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
