Ethereum’s roadmap now points to Dencun’s blobs, a proposed PeerDAS expansion and account abstraction to make rollups cheaper and applications more capable, but those advances will not automatically make a fragmented multi-chain experience simpler for users.
For years, Ethereum’s scaling debate was dominated by ambitious targets. Developers discussed rollups, data availability, modular blockchains and cryptographic proofs that could process activity away from the main chain while preserving Ethereum’s security. Those ideas have moved from research papers into production. The challenge now is less about proving that a new architecture is possible and more about making it work consistently for users, applications and businesses.
That shift is bringing network efficiency and transaction costs back to the center of Ethereum’s development agenda. Recent upgrades have lowered the cost of publishing rollup data, improved validator operations and introduced new tools for applications to sponsor or simplify transactions. At the same time, the ecosystem remains fragmented. Users may encounter different fees, wallets, bridges, tokens and liquidity pools depending on which layer-2 network they use. A transaction that is inexpensive on one network can still feel complicated, risky or surprisingly costly to someone moving between ecosystems.
The stakes extend beyond user convenience. Ethereum’s future as a settlement layer depends on whether rollups can grow without turning the base chain into a bottleneck or making the broader ecosystem difficult to navigate. Lower data costs can help decentralized exchanges, lending markets, payments applications and games. Better account infrastructure can make blockchain applications feel more like conventional software. More efficient validation can support a larger and more resilient network.
But the benefits are not automatic. Capacity upgrades can be unevenly distributed, applications may fail to pass savings to users, and competition is intensifying among alternative smart-contract platforms and specialized chains. Ethereum’s next phase will therefore be measured not only in transactions per second, but in the quality of the products built on top of it.
From expensive blocks to abundant data
Ethereum’s principal scaling strategy is built around rollups. These networks execute transactions outside the base layer, bundle the resulting activity and submit information to Ethereum for settlement or verification. The approach allows applications to serve more users without requiring every transaction to be processed independently by every Ethereum validator.
Rollups come in different forms. Optimistic rollups assume submitted transactions are valid unless challenged during a dispute period. Zero-knowledge rollups use cryptographic proofs to demonstrate that a batch was processed correctly. The engineering details differ, but both models require access to transaction data. That data allows users or other participants to reconstruct the rollup’s state and verify that the system is operating according to its rules.
Before Ethereum’s data-oriented upgrades, rollups generally published information through ordinary transaction calldata. Calldata competes with other transaction data for block space and was not designed specifically for the high-volume requirements of layer-2 networks. When Ethereum became busy, the cost of posting rollup data rose, and those expenses were passed through to users.
The Dencun upgrade, activated on March 13, 2024, introduced “blobs,” a separate temporary data structure designed to make rollup data cheaper. Blobs are not permanently stored in the same way as ordinary contract data. They remain available long enough for the network’s verification and recovery requirements, after which they can be pruned by nodes. This distinction reduces the storage burden while giving rollups a purpose-built market for data availability.
The immediate result was measurable on major networks, not merely theoretical. According to a March 20, 2024, CoinDesk analysis of Dune Analytics data, the average fee for a simple transaction on Base fell from about $0.31 before Dencun to roughly $0.0005 afterward. Optimism’s average fell from approximately $0.52 to $0.001, while Arbitrum’s declined from about $0.37 to $0.005. In other words, the data-cost reduction translated into roughly 99% lower fees on Base and Optimism and about a 99% decline on Arbitrum, although actual costs continued to vary with network demand and transaction type. Transfers and simple application interactions became far cheaper on several networks, and rollup operators gained room to offer lower fees or absorb more activity. For Ethereum, the change also represented a philosophical shift: the base layer was no longer trying to execute every transaction, but it was adapting its block structure to support a growing market for data publication.
That market remains a work in progress. Blob capacity is limited, and fees can rise when multiple rollups compete for available space. The objective of subsequent development is to increase the amount of data Ethereum can make available without imposing excessive bandwidth, storage or computational requirements on validators.
This is where proposals such as PeerDAS, or peer data availability sampling, enter the discussion. In broad terms, data availability sampling is intended to let nodes verify that large amounts of data are available without downloading every piece of it. A validator can sample portions of the data and use the network’s coding and sampling mechanisms to gain confidence that the full dataset could be recovered if needed.
The technical benefit is straightforward: Ethereum may be able to support more blob data while limiting the amount each individual node must handle. The economic implications are broader. If rollups can publish more data at lower and more predictable prices, they can increase capacity, serve larger applications and compete more effectively with monolithic blockchains that process activity directly.
However, more capacity does not automatically translate into better user economics. Rollup operators determine how fees are structured, how much margin they retain and whether they pass lower data costs to users. Some already charge very low fees for basic activity, while others may continue to price according to demand, execution costs or business strategy. Ethereum’s upgrades create the possibility of cheaper transactions; they do not guarantee that every user will see the full savings.
The cost of a transaction is more than gas
For users, “Ethereum fees” are often treated as a single problem, but the cost of an on-chain action has several components. On the base layer, users pay for computation and data through gas. On a rollup, they pay for execution on that network and for the cost of publishing data back to Ethereum. They may also encounter costs for bridging assets, swapping between tokens, signing multiple transactions or retrying a failed interaction.
This complexity is one reason that technical improvements can be difficult to recognize outside developer circles. A rollup may become substantially more efficient, yet a new user can still face a confusing interface that displays several balances, asks for a network switch and requires the user to acquire a specific token for fees. A payment is not truly frictionless if the user must first understand the architecture behind it.
Ethereum developers are therefore examining the user account itself. Traditional Ethereum accounts are controlled by private keys and require the user to hold the network’s native token to pay gas. This model is powerful but awkward for mainstream applications. Someone using a game, a social platform or a financial service should not necessarily need to purchase and manage a separate asset before completing a basic action.
Account abstraction offers a different model. Smart-contract-based accounts can support programmable permissions, recovery options, spending limits, batch transactions and sponsored fees. An application could pay the transaction cost on behalf of a customer, or a user could pay in a stablecoin rather than holding the network’s native token. Multiple actions could be bundled into one approval, reducing the number of prompts and opportunities for mistakes.
Recent Ethereum development has made these capabilities more accessible, including standardized transaction mechanisms that allow applications and specialized infrastructure providers to support smart accounts. The importance of this work is commercial as much as technical. A wallet that can recover access, limit risk and hide unnecessary network details is more likely to be used by people who are not already comfortable with blockchain systems.
There are trade-offs. Sponsored transactions require someone to finance the gas, and applications must decide whether the cost is justified by customer acquisition or engagement. Smart accounts can introduce new security assumptions, especially when they rely on third-party services for transaction routing or recovery. More flexible permissions may also create new attack surfaces if users do not understand what they have authorized.
Still, account-level improvements could be as consequential as cheaper block space. The history of technology shows that infrastructure adoption accelerates when complexity is moved away from the end user. Web browsers hid the details of network protocols; mobile operating systems standardized hardware access; payment applications turned banking instructions into a few taps. Ethereum’s account model is beginning to move in a similar direction.
Layer-2 growth creates a coordination problem
Ethereum’s rollup-centric strategy has produced an expanding collection of networks, each with its own applications, fee markets and operational choices. This diversity has advantages. Teams can optimize for gaming, trading, payments, privacy or institutional requirements. Competition among rollups can produce better tooling and specialized products. A failure in one network does not necessarily halt the entire ecosystem.
The same diversity can make Ethereum feel less like one network and more like a group of related environments. Assets may exist in canonical and bridged forms. Liquidity may be split across several decentralized exchanges. A user who moves funds to an unfamiliar network must consider whether a bridge is secure, whether an application is legitimate and whether the asset received will be accepted elsewhere.
For developers, fragmentation affects distribution. An application launched on one rollup may have access to a particular user base but limited liquidity or composability with contracts on another. Teams must decide whether to deploy on several networks, use interoperability tools or wait for standards that allow applications to function across chains. Every additional deployment can increase maintenance costs and enlarge the surface area for bugs.
The industry has responded with bridges, interoperability protocols and intent-based systems. In an intent model, users specify the outcome they want, such as exchanging one asset for another or moving funds to a target network, while specialized agents or “solvers” handle the route. This can hide some of the technical complexity, but it also introduces questions about pricing, execution quality, counterparty risk and transparency.
Ethereum’s own roadmap is increasingly concerned with making multiple execution environments feel connected without requiring the base layer to process every action. Improvements in data availability can support more rollups, while standardized account and transaction infrastructure can make those rollups easier to use. Yet the ecosystem still needs better conventions for addresses, asset representation, messaging and liquidity.
The business consequences are significant. Applications that solve cross-network complexity could become gateways into the broader Ethereum economy. Wallets, exchanges, payment providers and developer platforms are competing to control that gateway. The winning products may not be the ones that expose the most technical functionality, but those that reduce the number of decisions a user must make.
DeFi is an efficiency test
Decentralized finance provides one of the clearest tests of Ethereum’s infrastructure. Trading, lending, derivatives and asset management require frequent transactions, reliable settlement and deep liquidity. High fees can make small trades uneconomic, while delays and fragmented markets can create price differences that sophisticated traders exploit but ordinary users struggle to navigate.
Lower rollup costs can support a wider range of financial products. A decentralized exchange can process smaller orders without consuming a large share of a trader’s funds. Lending applications can make more frequent collateral adjustments. Stablecoin issuers can offer lower-cost transfers for remittances, payroll or business payments. Automated strategies can rebalance positions more often when the cost of doing so is predictable.
The opportunity is not limited to retail users. Financial institutions exploring tokenized funds, bonds or deposits need infrastructure that can support controlled access, transaction monitoring and predictable operational expenses. They may not require every product to settle directly on Ethereum’s base layer, but they may value an architecture in which multiple execution networks inherit security or settlement guarantees from a widely used public chain.
Efficiency also affects risk management. In volatile markets, a user may need to add collateral or close a position quickly. A fee spike or congested network can turn a manageable position into a liquidation. Faster and cheaper execution does not remove market risk, but it can reduce the chance that infrastructure costs determine the outcome.
At the same time, scaling can make financial applications more complex. New rollups may have different security stages, upgrade mechanisms and withdrawal assumptions. A protocol deployed across several networks may need to maintain separate liquidity pools and risk parameters. If an application treats all environments as equivalent when they are not, users may face hidden technical exposure.
For this reason, the next generation of DeFi infrastructure is likely to compete on operational quality rather than simply on the number of supported chains. Risk systems, monitoring, automated rebalancing and clear disclosure will matter as much as smart-contract design. A lower transaction fee is valuable, but a reliable system that explains where assets are held and how they can be recovered is more valuable still.
Builders are watching predictability, not just the lowest fee
Developers often care less about the absolute minimum transaction cost than about whether costs can be forecast. An application that occasionally pays almost nothing but experiences severe fee spikes may be harder to operate than one with a slightly higher but stable cost structure.
Predictability matters for product design. A consumer application may want to subsidize user transactions, but it cannot do so responsibly if its expenses are impossible to estimate. A trading protocol needs to know whether its market-making strategy will remain profitable when data demand rises. A gaming company must decide whether on-chain actions can be included in a subscription or advertising model.
Ethereum’s blob market and future data-capacity improvements are intended to make rollup economics more scalable, but developers will continue to track how costs behave during demand surges. They will also evaluate the operational requirements of running nodes, sequencing transactions and proving batches. The cheapest theoretical architecture is not necessarily the most attractive if it is difficult to monitor or dependent on a small number of infrastructure providers.
Sequencing is a particularly important issue. Many rollups use a centralized sequencer to order transactions, which improves speed and user experience but creates a potential point of failure or control. Decentralized sequencing can improve resilience, though it may involve more complicated coordination and new economic mechanisms. Ethereum’s base layer cannot solve every rollup design choice, but its standards and security model can influence which models become viable.
Proof generation is another area where efficiency can determine adoption. Zero-knowledge systems can compress complex computation into proofs that Ethereum verifies relatively cheaply. Generating those proofs, however, can require specialized hardware and significant engineering. Improvements in proving algorithms, software and hardware could reduce costs and make more applications suitable for zero-knowledge execution.
The commercial ecosystem around these tools is expanding. Cloud providers, node operators, proving services, wallet companies and developer platforms are turning protocol improvements into products. This creates a healthier division of labor, allowing application teams to use advanced infrastructure without building every component themselves. It also introduces concentration risk if too much activity depends on a small group of service providers.
For Ethereum’s roadmap to translate into durable growth, developers will need choices. A robust ecosystem should support multiple clients, proving systems, sequencers and infrastructure providers. Diversity is less visible than a headline throughput number, but it is central to resilience.
Validators remain part of the efficiency equation
Scaling discussions can focus heavily on users and rollups while overlooking the validators that secure Ethereum. Every increase in capacity has consequences for bandwidth, storage, computation and hardware requirements. If operating a validator becomes too expensive or technically demanding, participation could become concentrated among large operators.
Ethereum’s proof-of-stake model depends on a broad set of participants. The network does not need every participant to have identical resources, but it must avoid a situation in which only specialized institutions can reliably validate the chain. Developer work on client efficiency, block propagation and data handling therefore serves a security objective as well as a performance objective.
The challenge is balancing capacity with accessibility. Larger blocks or more blob data can support more activity, but nodes must receive and process that information within tight time limits. Data availability sampling and improved networking are designed to loosen this constraint, allowing the protocol to scale without requiring every node to store or download everything permanently.
This is a difficult engineering boundary. Changes that look efficient from a transaction perspective may add complexity for clients or increase the risk of implementation bugs. Ethereum’s conservative development culture reflects the value placed on avoiding consensus failures. Upgrades are tested across multiple clients, examined by researchers and deployed only after extended coordination.
That process can appear slow compared with newer networks that advertise rapid iteration. But Ethereum’s settlement role gives delays a different meaning. A mistake in a widely used base layer can affect exchanges, stablecoins, lending markets, rollups and custody providers simultaneously. The cost of an upgrade is therefore measured not only in development time, but in the risk of disrupting a large economic system.
The next milestones will be judged against this standard. Increasing capacity is useful only if the network remains secure, decentralized and operationally understandable. Ethereum’s developers are effectively trying to improve the road while ensuring that the road continues to support heavy commercial traffic.
Competition is forcing the issue
Ethereum is not improving in isolation. Rival smart-contract platforms continue to compete on transaction speed, cost, user experience and application incentives. Some promote high throughput on a single chain. Others use alternative modular designs, specialized virtual machines or integrated exchanges and wallets. Developers have more options than they did during Ethereum’s early expansion.
The competitive threat is not simply that users will choose a cheaper network for one transaction. Developers may choose a different platform if they believe it offers a more coherent environment, stronger consumer tooling or a clearer path to scale. Liquidity and network effects remain powerful, but they are not permanent advantages.
Ethereum’s response is to combine a strong settlement and security layer with a broad collection of execution environments. That model can work if the underlying network provides valuable guarantees and if users do not experience the ecosystem as a maze. If cross-network movement is expensive or confusing, the benefits of shared settlement may be difficult to communicate.
This is why efficiency has become a product issue. A blockchain can have excellent cryptography and an active developer community, but adoption depends on whether applications deliver useful outcomes at an acceptable cost. A stablecoin transfer for a freelancer, a digital asset purchase for a consumer or a tokenized fund subscription for an institution must be simple enough to fit into an existing workflow.
Ethereum’s large developer base gives it an important advantage in this competition. Tooling, standards and composable protocols can reduce the cost of launching new products. But that advantage must be converted into experiences that users recognize as better. Infrastructure is successful when it disappears into the product.
What the next milestones will reveal
Investors and application builders are likely to monitor several indicators as Ethereum’s development cycle advances.
The first is the growth and behavior of blob demand. If rollups use more blob space while fees remain manageable, it would suggest that Ethereum is expanding capacity effectively. If demand quickly overwhelms supply, developers may face pressure to accelerate data-related improvements or reconsider how capacity is allocated.
The second is whether lower infrastructure costs appear in end-user pricing. Rollup transaction fees, stablecoin transfer costs and decentralized exchange execution will provide a practical measure of how much value is reaching customers. A reduction in operator expenses that does not improve user economics may still benefit the ecosystem, but it will be less persuasive as an adoption story.
The third is application distribution across networks. More deployments do not necessarily indicate progress if liquidity and users remain thinly spread. Builders will look for evidence that wallets, bridges and interoperability systems are making multi-network applications easier to use rather than simply adding more destinations.
The fourth is the adoption of smart accounts and sponsored transactions. These tools could be a turning point if they become standard components of consumer applications. Their success will depend on security, wallet support, clear permissions and sustainable sponsorship models.
The fifth is validator and client diversity. Higher capacity must not come at the expense of network resilience. Researchers and operators will continue to assess whether upgrades increase hardware requirements or create new concentration points in sequencing, proving and data services.
Finally, businesses will watch whether Ethereum can support use cases beyond speculative trading. Payments, tokenized financial products, decentralized physical infrastructure, gaming and social applications all impose different demands. A network that can serve these categories at predictable costs will have a stronger claim to being foundational digital infrastructure.
The measure of progress is a better default experience
Ethereum’s technical roadmap is often described through specialized terms: blobs, rollups, data availability sampling, account abstraction and zero-knowledge proofs. Each represents a significant engineering effort, but none is the final product. The final product is the experience that users and companies receive when they interact with an application.
That experience should include fees that are understandable, settlement that is dependable and accounts that do not require users to become security experts. It should allow developers to scale without rebuilding their application for every network. It should give institutions enough operational clarity to integrate blockchain systems into existing financial and commercial processes.
Achieving that outcome will require more than base-layer upgrades. Wallets must improve their interfaces. Rollups must communicate their security and withdrawal models. Bridges and interoperability protocols must become safer. Applications must decide when to abstract complexity and when to disclose it. Businesses must build sustainable fee and sponsorship models rather than relying indefinitely on incentives.
Ethereum’s developers can provide the underlying capacity and standards, but the ecosystem’s companies and builders will determine whether those capabilities become useful products. This is a familiar pattern in technology: infrastructure creates the possibility, while product design determines adoption.
The current focus on efficiency is therefore not a retreat from Ethereum’s larger ambitions. It is an attempt to make those ambitions practical. A secure settlement network with abundant data and flexible accounts could support a much broader digital economy than one optimized primarily for high-value transactions by experienced crypto users.
The next phase will test whether Ethereum can turn its technical lead into an everyday advantage. If it succeeds, users may stop thinking about which chain they are on, how a rollup posts data or where a transaction is settled. They will simply experience applications that are faster, cheaper and easier to use. That quiet disappearance of infrastructure would be one of the clearest signs that the network’s development strategy is working.