Ethereum’s Glamsterdam upgrade is shaping up to be the network’s most consequential architectural change since The Merge — bringing parallel transaction execution to the base layer and targeting a gas limit of up to 200 million, more than triple today’s levels.
By Amir Hassan | June 19, 2026
The Architecture
For years, Ethereum’s scaling strategy relied almost entirely on layer-2 networks like Arbitrum, Base, and Optimism. These rollups absorbed user activity while Ethereum’s base layer settled into a role as a coordination and settlement hub. The approach worked — Ethereum’s layer-2 ecosystem now processes over 105 million daily transactions across all rollups, while the mainnet itself handles just over 2 million per day, according to data from L2Beat.
But this created a problem: fragmentation. Liquidity, users, and applications scattered across dozens of chains. The base layer itself remained constrained at roughly 36 transactions per second, a ceiling that has persisted even as rival networks demonstrated higher throughput.
Glamsterdam changes the architecture itself. The headline innovation is parallel execution — the ability for the network to process multiple non-conflicting transactions simultaneously across different CPU cores, rather than one at a time in sequence. Think of it like a grocery store opening additional checkout lanes: instead of every transaction waiting in a single line, independent transactions can be processed in parallel.
This is enabled by EIP-7928, which introduces what developers call “BALs” (Block Access Lists). BALs reorganize how the network records data dependencies, allowing nodes to pre-fetch data and safely execute transactions in parallel. Alongside this, gas repricing proposals like EIP-7904 and EIP-8037 modernize Ethereum’s fee model by aligning computational costs with current hardware realities — replacing assumptions that were baked in years ago.
Consensus Mechanisms
Glamsterdam also introduces changes to how blocks are produced. A key proposal is the implementation of ePBS (enshrined Proposer-Builder Separation), which restructures the relationship between validators who propose blocks and the entities that build them.
In plain terms, ePBS separates the roles of “deciding which transactions go in a block” from “actually constructing the block.” This reduces the influence of MEV (maximal extractable value) extraction on regular users and makes the network more resistant to censorship attempts. Validators retain their role in consensus — staking ETH to secure the network — but the mechanics of block construction become more transparent and decentralized.
The parallel execution layer does not change Ethereum’s proof-of-stake foundation. ETH holders who stake their tokens continue to earn rewards for validating transactions. What changes is how efficiently those validators can process the transactions they receive.
Network Health
The timing of Glamsterdam matters. Ethereum is currently navigating a difficult market period. ETH trades around $1,702 as of this writing, down significantly from its November 2025 high near $4,955. The broader crypto market remains pressured by macroeconomic factors including persistent inflation and limited monetary easing from central banks.
Yet beneath the price weakness, network fundamentals tell a different story. Daily active addresses remain robust. Layer-2 transaction volume continues to grow. And with Bitcoin at approximately $63,132 and Solana at $69.26, the broader market is treading water — waiting for a catalyst.
Historically, Ethereum upgrades have served as exactly that kind of catalyst. ETH gained roughly 77% in the months before The Merge in 2022, about 55% ahead of Shanghai in 2023, 76% before Dencun in 2024, and nearly 89% before Pectra in 2025, according to market data compiled by MarketScreener. Of course, past performance does not guarantee future results — and these upgrades also tended to trigger “sell the news” reactions once deployed.
Developer Ecosystem
Testing for Glamsterdam is actively underway across Ethereum testnets. The upgrade follows Ethereum’s established pattern of incremental, well-tested protocol changes rather than sweeping overhauls. After Glamsterdam, developers have already mapped out the next upgrade — Hegota, tentatively planned for late 2026 — continuing a faster release cadence.
For developers building on Ethereum, parallel execution opens new possibilities. Applications that were previously too computationally expensive for layer-1 — complex DeFi operations, high-frequency trading, and data-heavy smart contracts — may become feasible directly on mainnet. This could reduce reliance on layer-2s for certain use cases while keeping rollups relevant for others.
The gas limit expansion is particularly significant for builders. Ethereum’s gas limit hovered near 30 million for years before rising toward 60 million in early 2025. Glamsterdam targets a floor of up to 200 million — a dramatic increase that would allow far more computation per block. Developers estimate this could push layer-1 throughput toward 100 transactions per second, up from the current 36 TPS.
Final Assessment
Glamsterdam represents Ethereum’s most direct answer to the scaling question that has dominated crypto discourse for years. Where Solana and other high-throughput chains achieved speed through more centralized validator structures or heavier hardware requirements, Ethereum is attempting to scale without sacrificing decentralization or node accessibility.
For everyday investors, the upgrade matters in two ways. First, it could reduce transaction costs and congestion on Ethereum’s base layer — the network you interact with directly when you transfer ETH or interact with mainnet smart contracts. Second, higher throughput and greater efficiency could support increased ETH burning under EIP-1559 during high-activity periods, which reduces supply over time.
The mainnet is not getting cheaper to use overnight. Layer-2 networks will continue handling most user activity for the foreseeable future. But Glamsterdam signals that Ethereum’s base layer is no longer standing still — it is actively evolving to meet the demands of an ecosystem that processes over a hundred million transactions daily across all its chains combined.
As Bankless founder Ryan Sean Adams put it: “Ethereum L1 just woke up.”
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
parallel execution on L1 is huge but 200M gas limit gonna make node requirements insane. wonder how many validators drop off
been saying this for months, L2 fragmentation was killing ETH. fixing base layer is the only real path
^ fixing L1 means L2s lose their reason to exist though right? optimistic rollups entire value prop was cheaper fees
not really, L2s still cheaper even with 200M gas. its about orders of magnitude not absolute throughput
Rune H. L2s dont die they just become cheap execution sandboxes. glamsterdam makes L1 competitive for settlement-heavy apps while L2s handle the high-frequency stuff
Katrin the L2s wont die, they just become execution environments while L1 handles settlement + data availability. different roles not extinction
105M daily txs on L2s vs 2M on mainnet and they finally realized the fragmentation problem. better late than never i guess
105M daily tx on rollups vs 2M on mainnet. the gap is so wide already that even tripling base layer doesnt catch up
200M gas limit with parallel execution is huge. thats not a tweak, thats a fundamental rethink of the EVM. real question is validator hardware requirements
yeah this is gonna price out home stakers for sure. 200M gas means bigger machines, more storage, better bandwidth. say goodbye to decentralization
solo_staker_ already running on fumes. 200M gas means a new machine which means solo staking becomes a rich persons game. say goodbye to the last pretense of decentralization
200M gas limit with parallel EVM is the real unlock. tripling throughput on L1 means L2s actually have to compete on price instead of existing by default
36 to 100 tps on L1 is nice and all but its still 105m vs 2m daily. the gap is so wide tripling barely closes it
105M daily tx on rollups vs 2M on mainnet and people think tripling L1 throughput fixes anything? the gap is 50x not 3x
stake_threshold_ tripling L1 doesnt close the 50x gap but it makes L1 competitive for settlement heavy apps that cant afford L2 bridge risk. different use case
l1_maximalist_ bridge risk is the real driver. 3 bridging exploits this year and people still ask why L1 throughput matters. settle on mainnet or accept the risk
BALs in EIP-7928 are genuinely clever engineering. pre-fetching access lists so nodes can parallelize without conflicts, never seen that approach on evm before
BALs in EIP-7928 letting nodes prefetch access lists is the kind of deep engineering that actually makes parallel execution viable. not just slapping threads on the EVM
200M gas limit with parallel execution means validators need serious hardware. expecting a 15-20% validator drop in the first quarter post-upgrade
Heike S. 15-20 percent validator drop would actually improve decentralization metrics since the dropoff hits data centers first not home stakers. ironic
Heike S. a 15-20 percent validator drop improving decentralization metrics because data centers leave first is peak unintended consequences. love it
200M gas limit with parallel execution is the real deal. current 30M limit means every congested block is serialized. parallelization alone could 3x throughput without touching L2s
parallel_evm_maxi_ agreed but 200M gas limit means state bloat explodes. who runs a full node when the state size hits multiple TB? the decentralization tradeoff is real
105M daily tx on L2s vs 2M on mainnet. the base layer is becoming a settlement hub not a execution platform. Glamsterdam trying to reverse that trend
200M gas limit sounds great until you realize the sync requirements for new nodes. try bootstrapping a full archive node after this upgrade