Ethereum Glamsterdam Upgrade: How Block Access Lists and ePBS Will Transform the Network in 2026
Ethereum validators currently process transactions the way a grocery store checkout works with a single lane: one item at a time, in order, no matter how long the line stretches. The Glamsterdam upgrade, scheduled for mid-2026, fundamentally changes this architecture. By introducing Block Access Lists (BAL) and enshrined Proposer-Builder Separation (ePBS), Ethereum is preparing to scale from roughly 21 transactions per second to 10,000 TPS—a 476x improvement that could reshape DeFi, NFTs, and on-chain applications.
The Bottleneck Problem: Why Ethereum Needs Parallel Execution
Since the Merge in 2022, Ethereum has operated on proof-of-stake consensus, but its execution layer has remained fundamentally sequential. Every transaction must wait for the one before it to complete, creating a computational traffic jam during high-demand periods. This design choice made sense when Ethereum was younger—sequential execution is simpler to implement and reason about—but it now represents the network's most significant scaling limitation.
The numbers tell the story clearly. Ethereum's base layer currently handles approximately 21 TPS at a 60 million gas limit. Meanwhile, competitors like Solana process 1,100+ TPS, and centralized payment networks like Visa handle over 65,000 TPS. For Ethereum to serve as the settlement layer for global finance—BlackRock's BUIDL fund, JPMorgan's tokenized money market funds, and the growing institutional DeFi ecosystem—it needs dramatically higher throughput.
Layer 2 solutions like Arbitrum, Optimism, and Base have absorbed much of this demand, collectively holding over $39 billion in TVL. But L1 congestion still creates problems: expensive rollup posting costs, delayed finality, and the persistent question of whether Ethereum can remain relevant as L2s capture more value.
Block Access Lists: From Single Lane to Multi-Lane Highway
The Glamsterdam upgrade's centerpiece is EIP-7928, which introduces Block Access Lists. Despite the bureaucratic-sounding name, BALs enable something transformative: they create a "map" showing how transactions within a block relate to each other and which parts of the network's state they modify.
Here's why this matters. When two transactions access completely different accounts and storage slots—say, one user swapping ETH for USDC on Uniswap while another mints an NFT—there's no logical reason they can't execute simultaneously. The current architecture forces them to wait in line anyway, because the Ethereum Virtual Machine doesn't know in advance whether they might conflict.
Block Access Lists solve this by declaring upfront exactly which accounts and storage locations each transaction will touch. Historical data analysis reveals that 60-80% of transactions in a typical block access completely disjoint storage slots. These transactions can execute in parallel immediately. The remaining 20-40% with potential dependencies can be parallelized using post-transaction state diffs that track intermediate changes.
Gabriel Trintinalia, a senior blockchain engineer at Consensys, describes the improvement succinctly: the system eliminates Ethereum's biggest bottleneck by "preloading necessary data into memory rather than sequentially reading from disk." Instead of waiting to discover what each transaction needs, validators can prepare all required state data the moment they see a block's access list.
The practical result: Glamsterdam targets a gas limit increase from 60 million to 200 million per block—a 233% jump—combined with parallel processing that could push throughput past 10,000 TPS. For users, this translates to lower fees and faster confirmations. For developers, it means building applications without worrying about network capacity constraints.
Enshrined Proposer-Builder Separation: Protocol-Level MEV Management
The second major component of Glamsterdam is EIP-7732, which brings Proposer-Builder Separation directly into Ethereum's consensus rules. To understand why this matters, you need to understand how blocks get built today.
Currently, validators can either build their own blocks or outsource this work to specialized "builders" through MEV-Boost, an off-chain system developed by Flashbots. Builders compete to extract Maximal Extractable Value—profits from transaction ordering, sandwich attacks, and arbitrage—and share some of that value with validators in exchange for block inclusion rights.
The problem? This system depends on trusted intermediaries called relays, creates centralization risks, and operates entirely outside Ethereum's protocol rules. A small number of sophisticated builders dominate the market, capturing value that could otherwise flow to the broader validator set.
Enshrined PBS moves this entire process on-chain, replacing trusted relays with protocol-enforced commitments. Under ePBS, builders submit sealed bids and block header commitments directly to the network. Validators—who may be running on home hardware without specialized MEV infrastructure—simply select the highest bid. The winning builder must then reveal the full block contents within a specified window, verified by a new Payload Timeliness Committee.
This architecture delivers several benefits. Solo stakers gain access to MEV rewards without running complex builder software. The protocol itself enforces builder commitments, reducing trust requirements. And standardized rules make it harder for any single entity to monopolize block production.
However, challenges remain. Academic research estimates that builders might strategically delay block revelation on approximately 0.82% of blocks under normal conditions, rising to 6% during high-volatility periods—a phenomenon called the "free option problem." If market conditions change significantly during the revelation window, builders might benefit from withholding their blocks. The current ePBS specification doesn't fully address these edge cases, though developers continue refining the design.
The 2026 Ethereum Roadmap: Glamsterdam in Context
Glamsterdam doesn't exist in isolation. It follows two significant 2025 upgrades that laid the groundwork for parallel execution:
Pectra (May 2025) raised Ethereum's blob throughput, enabling cheaper data availability for rollups. It also introduced account abstraction improvements through EIP-7702, allowing externally owned accounts to temporarily behave like smart contract wallets—enabling features like gas sponsorship and batched transactions that improve user experience.
Fusaka (December 2025) activated PeerDAS, expanding data availability sampling from 6 to 48 blobs per block. This upgrade specifically targets rollup efficiency, reducing costs for the L2 ecosystem while maintaining Ethereum's security guarantees.
Glamsterdam, tentatively scheduled for Q2/Q3 2026, builds on this foundation. Community documentation references a June 2026 target, though developers emphasize this remains aspirational pending validation of critical components like the BAL devnet (which recently came online with the Reth client working successfully).
Looking further ahead, Ethereum has already named its next upgrade: Hegota in H2 2026. This upgrade will focus on privacy and state management through Verkle Trees—a cryptographic structure that dramatically reduces the data validators need to store while enabling stateless clients. Combined with Glamsterdam's throughput improvements, Hegota could complete "The Surge" portion of Ethereum's roadmap, pushing the ecosystem toward 100,000+ TPS across L1 and L2 combined.
What 10,000 TPS Means for the Ethereum Ecosystem
Ethereum's scaling roadmap isn't just about bigger numbers—it's about enabling new categories of applications that aren't viable today.
DeFi Protocols currently make tradeoffs around gas costs. Aave V4, launching in early 2026, introduces a hub-and-spoke architecture to unify liquidity across networks partly because fragmenting across L2s has become operationally complex. With 10,000 TPS on L1, protocols could deploy directly on mainnet without worrying about congestion during high-volatility periods.
Institutional Adoption depends on predictable costs and settlement times. JPMorgan's tokenized money market fund and BlackRock's BUIDL operate in a world where transaction failures or multi-minute delays are unacceptable. Higher throughput and lower fees make Ethereum more credible as settlement infrastructure for traditional finance.
Real-World Assets represent a $358 billion market that's rapidly tokenizing. Private credit, US treasuries, and commodities all require high-throughput blockchains that can handle the volume of a functioning financial system. Ethereum's credibility with regulators and institutions makes it a natural home for RWAs—if it can scale.
NFTs and Gaming largely migrated to L2s due to mainnet costs. Parallel execution could bring some activity back, particularly for applications that benefit from Ethereum's composability and security guarantees.
Risks and Tradeoffs
No upgrade of this magnitude comes without concerns. The parallel execution model requires careful handling of edge cases where transactions do have dependencies. If the access list mechanism has bugs, it could enable double-spend attacks or other security issues. The BAL devnet is still early, and significant testing remains before mainnet deployment.
ePBS introduces new attack vectors around builder behavior. The "free option problem" gives builders economic incentive to game the system during volatile periods. While the protocol includes fallback mechanisms to maintain liveness, the long-term dynamics of an ePBS market are unknown.
There's also the question of centralization. Even with ePBS making MEV rewards more accessible, capital-intensive builders with sophisticated infrastructure might still dominate block construction. The upgrade improves the current situation but doesn't fully solve MEV-related centralization concerns.
The Competitive Landscape
Ethereum isn't upgrading in a vacuum. Solana continues capturing market share, particularly in memecoin trading and AI agent applications—processing $118 billion in DEX volume over a recent 30-day period compared to Ethereum's $40 billion. Alternative L1s like Sei Network emphasize parallel execution as a core feature, while Cosmos-based chains and newer entrants compete for specific use cases.
The Glamsterdam upgrade represents Ethereum's answer to these competitive pressures. By achieving 10,000 TPS on L1 while maintaining decentralization, the network aims to prove that it can scale without sacrificing its core values. Whether this roadmap executes on time—and whether users care about decentralization enough to prefer Ethereum over faster alternatives—remains to be seen.
Looking Ahead
Ethereum's 2026 roadmap is ambitious. Block Access Lists and enshrined PBS represent fundamental changes to how the network operates, not incremental improvements. Success would cement Ethereum's position as the settlement layer for tokenized finance. Delays or bugs could push more activity to competing chains.
For developers, the message is clear: Ethereum is betting heavily on L1 scalability. Applications designed around L2 constraints may need to reconsider their architecture as mainnet becomes viable for high-throughput use cases.
For investors and users, Glamsterdam represents a potential inflection point. Ethereum's value proposition has always been security and decentralization over raw speed. If the network can deliver both, the narrative around "Ethereum can't scale" loses much of its force.
The BAL devnet is live. The specifications are public. The timeline is set. Now comes the hardest part: shipping.
For developers building on Ethereum and other leading chains, BlockEden.xyz provides enterprise-grade RPC infrastructure designed to support high-throughput applications. Explore our API marketplace to access reliable node services across the ecosystem.