Author: YQ

Compiled by: Yangz, Techub News

Shift from Philosophy to Infrastructure

Fundamental Change: In 2015, Ethereum embraced decentralization, censorship resistance, and trustlessness as its ultimate values (the goal itself, rooted in the cypherpunk ideology). By 2025, these characteristics have shifted to serve instrumental value (as means to achieve practical goals, such as eliminating counterparty risk, achieving global financial inclusion, and reducing operational costs). This shift is crucial: idealism asks, "How decentralized can we make this?", while pragmatism asks, "How decentralized do we need to be to solve this problem?" The Devconnect conference in Buenos Aires in 2025 revealed that the Ethereum Foundation has decisively chosen to pursue the latter question.

When Ethereum launched in July 2015, its founders articulated a vision rooted in the 1990s cypherpunk movement. Just like the 2,000 mathematicians, cryptographers, and software engineers who resisted government attempts to ban encryption, Ethereum's early community pursued decentralization and censorship resistance as ultimate values. The project was to operate as a "world computer," where smart contracts executed with mathematical certainty, free from corporate or governmental interference. This philosophical commitment supported the project through its early years, attracting developers more for ideological reasons than for commercial interests.

A decade later, a fundamentally different Ethereum was presented at the five-day Devconnect 2025 technical conference in Buenos Aires. Presentations by Ethereum Foundation researchers and application builders revealed a decisive shift: from pursuing the perfection of abstract protocols to pragmatically improving infrastructure for practical applications. This transformation became evident in the structure of the event itself. Ethereum Day on November 17 kicked off with Tomasz Stanczak, Hsiao-Wei Wang, Ansgar Dietrichs, Barnabé Monnot, and Vitalik Buterin introducing the Ethereum Foundation's strategic restructuring around three specific priorities established in April 2025: scaling Layer 1 by increasing the Gas limit, expanding Blob data availability through the deployment of PeerDAS, and improving user experience through cross-chain interoperability. Specific achievements delivered in 2025 were then detailed: the Gas limit doubled from 30 million to 60 million in December (via the Fusaka upgrade), the number of validators exceeded 1.1 million, and $70 billion worth of staked ETH was used to secure the network. The following days focused on specific application areas rather than abstract protocol theory. Stani Kulechov elaborated on decentralized credit markets, and Santiago Palladino detailed Ethereum's interoperability layer. On November 19 (Trustless Agent Day), the ERC-8004 portable reputation protocol and the x402 micropayment protocol for autonomous AI agent economies were introduced. The Privacy Summit and ZK Day showcased that zero-knowledge proofs achieved an average block verification speed of under 10 seconds. Throughout, speakers emphasized solving concrete problems for actual users, such as cross-chain fragmentation, DeFi privacy needs, institutional settlement requirements, and autonomous agent payment infrastructure.

We witnessed Ethereum's evolution from idealism to pragmatism across five key infrastructure areas, where the ultimate philosophical goals have given way to instrumental engineering objectives: Layer 1 scaling achieved through incremental optimization rather than architectural revolution, cross-layer interoperability addressing Layer 2 fragmentation, DeFi primitives empowering a $300 trillion credit market, autonomous agent infrastructure aimed at a machine-native economy, and institutional adoption frameworks prioritizing privacy over transparency.

L1 Scaling: Incremental Optimization Replaces Architectural Revolution

Strategic Restructuring and 3x Annual Goals

In the opening speech of Ethereum Day, Ansgar Dietrichs and Barnabé Monnot detailed the restructuring undertaken by the Ethereum Foundation in April 2025. This marked a shift from years of research on sharding technology to a pragmatic path for throughput improvement achievable in the near term. The Foundation is no longer pursuing major architectural changes that could take 5 to 10 years but is committed to achieving a 3x annual throughput increase through systematic client optimizations and targeted protocol adjustments. This approach reflects the hard-earned lessons about the cost of complexity in distributed systems: coordinating between four independent execution clients (Geth, Nethermind, Besu, Erigon) and five consensus clients (Prysm, Lighthouse, Teku, Nimbus, Lodestar) makes significant protocol changes costly in terms of both development time and deployment risk.

The strategy for increasing the Gas limit will advance during the "Pragmatic Scaling Era" (2025-2026), addressing bottlenecks through incremental optimization rather than architectural revolution. The increase from 30 million to 60 million is achieved through client performance optimizations and the EIP-7623 re-pricing of call data (which charges 40 gas per byte for L2 Rollups with a high proportion of call data, compared to 16 gas per byte for standard transactions) as well as the EIP-7825 transaction Gas limit cap (16.78 million per transaction). The entire process is divided into three steps: increasing from 30 million to 36 million (February 2025), from 36 million to 45 million (July 2025), and from 45 million to 60 million (November 2025), with the default value officially set to 60 million in the Fusaka upgrade on December 3.

This change, combined with the introduction of EIP-4844 dedicated Blob transactions through the Dencun upgrade in March 2024, provides Rollups with an independent data availability layer while freeing up block space for L1 execution. Recent scaling efforts focus on: enhanced proposer-builder separation (ePBS), block-level access lists supporting parallel execution (BAL), targeted re-pricing that aligns Gas costs with actual computation costs, and doubling the block rate with a 6-second block interval. Longer-term sustainability planning (2027-2030) looks at streamlining consensus mechanisms, virtual machine replacement, binary tree state structures, and protocol simplification, rather than the previously planned Verkle tree scheme (which was abandoned due to quantum computing vulnerabilities in polynomial commitment schemes).

Client Performance Benchmarks and Engineering Constraints

The Fusaka upgrade has obtained precise benchmark data on the Sepolia testnet and the mainnet shadow fork. The Geth client, serving about 60% of the validator set, takes 3.0 seconds to process a block at the full 60 million Gas limit, achieving a throughput of 20 million Gas per second. Nethermind executes the fastest at 2.4 seconds (25 million Gas per second), while Besu takes 3.3 seconds (18 million Gas per second), and Erigon completes block processing in 2.7 seconds (22 million Gas per second). All implementations perform well below the critical threshold of 4 seconds, ensuring that 90% of validators can receive and process blocks within the first quarter of a 12-second slot, thus maintaining consensus security margins. Network propagation analysis shows that 90% of validators receive blocks within 0.7 to 1.0 seconds via the Gossip protocol, but the remaining 10% may experience delays of 2 to 3 seconds due to geographical differences. These engineering realities prompt a conservative incremental increase in the Gas limit rather than a sudden leap that could jeopardize network stability.

Currently, the bottleneck has shifted from raw execution speed to state access patterns, disk I/O, and cumulative state growth. Measurements indicate that for complex transactions, accessing state accounts and storage slots now dominates execution time. At a 60 million Gas limit, the annual growth rate of state data is approximately 60 GB. If expanded to 300 million Gas without mitigation, the annual growth rate would reach 300 GB, and the state size could reach several TB within a few years. This reality was originally a driving force behind Verkle tree research, but advancements in quantum computing have forced a shift to pragmatic near-term management strategies (active state pruning, state rental economic mechanisms), while developing quantum-resistant binary tree alternatives for the 2027-2030 period.

PeerDAS and Blob Expansion for L2 Rollups

The second strategic priority is to expand Blob data availability, directly addressing the needs of L2 Rollups. The current infrastructure supports 3 to 6 Blobs (each 128 kilobytes) per block, providing a capacity of 384 to 768 kilobytes in each 12-second slot. The PeerDAS (Peer-to-Peer Data Availability Sampling) deployed through the Fusaka upgrade, utilizing erasure coding mathematics, can scale to 16 Blobs in the short term and potentially to 64 Blobs in the long term.

Each Blob is split into multiple data fragments using Reed-Solomon coding, allowing any 50% of the fragments to reconstruct the complete data. Validators download random subsets rather than the entire Blob, with the network ensuring data availability through collective cooperation, requiring no single validator to store all content. This sampling method can expand the number of Blobs by 10 times or more while reducing the bandwidth required for each validator from O(n) to O(log n).

Deployment Timeline: Development network testing in Q3 2025, testnet in early 2026, and mainnet activation in mid-2026 (subject to security review). Once operational, PeerDAS can increase Rollup data availability by 10 times while lowering the Gas price of Blobs by expanding capacity.

ZK-EVM Proofs: From Theory to Production Timeline

Vitalik Buterin's discussion on agent infrastructure and Ansgar's protocol updates both highlighted significant breakthroughs in zero-knowledge Ethereum Virtual Machine (ZK-EVM) proof times. Multiple ZK-EVM teams achieved average proof times of under 10 seconds in 2025, a substantial reduction from the 5 to 10 minutes seen in 2024. This marks a critical step for Ethereum towards achieving real-time proofs (under 12 seconds to match slot times).

Deployment will follow an incremental path: the first phase will experimentally introduce validity proofs, allowing validators to choose to verify ZK proofs instead of re-executing parts of blocks; the second phase will implement a hybrid model where key blocks require ZK proofs while most blocks continue normal execution; the third phase will transition to a proof-mandatory model where all blocks must have ZK proofs; the fourth phase will realize a complete ZK-EVM, allowing stateless clients to operate without storing state, thus providing full security for mobile and browser nodes. As for the implementation timeline, production deployment is expected to be completed between 2027 and 2030.

Cross-Layer Interoperability: Addressing L2 Fragmentation

Fragmentation Issues

In his "Ethereum Everywhere" talk, Santiago Palladino revealed a fundamental contradiction in Ethereum's development roadmap centered around Rollups. Despite over 50 L2s achieving a combined throughput of more than 100,000 transactions per second, fragmentation has led to severe user experience issues and liquidity segmentation, threatening the value proposition of a unified Ethereum ecosystem. A user holding assets on Arbitrum cannot purchase an NFT on zkSync without bridging funds through L1, which requires a 7-day waiting period (the fraud proof window for optimistic Rollups) and a $35 Gas fee. Liquidity is fragmented across different chains, and the trading prices of the same token vary on different L2s. Applications must be deployed separately on each chain, dispersing developers' focus and user bases.

Ethereum Interoperability Layer: Single Signature for Multi-Chain Operations

The Ethereum Interoperability Layer (EIL) is collaboratively developed by teams from Arbitrum, Optimism, Polygon, zkSync, and Base, built on ERC-4337 account abstraction, aiming to achieve cross-chain operations through a single signature. Its technical innovation lies in Merkle tree batch authorization. Users construct an operation tree spanning multiple chains, sign the Merkle tree root, and submit branches to each target chain. Smart contract accounts on each chain verify the Merkle proof based on the signed tree root, enabling atomic multi-chain execution without complex cross-chain messaging protocols.

The mechanism demonstrated by Palladino at the conference showcased specific efficiency improvements: a user holding 10,000 USDC on Arbitrum wishing to purchase an NFT for 5,000 USDC on zkSync only needs to sign a single Merkle tree root, simultaneously authorizing the deduction on Arbitrum and the purchase on zkSync. Cross-chain liquidity providers (XLPs) can "pre-run" settlements by immediately providing 5,000 USDC on zkSync, then claiming the user's funds from Arbitrum after the withdrawal delay period. XLPs charge about $5 (0.1%) for this service, allowing the transaction to be completed within 1 minute from the user's perspective, compared to over 7 days and $35 with traditional bridging methods.

Based on Account Abstraction

ERC-4337 enables the EIL by replacing externally owned accounts (controlled by ECDSA private keys) with programmable smart contract accounts. Traditional Ethereum addresses can authorize only one operation per signature. Smart contract accounts can implement arbitrary verification logic, including verifying Merkle proofs that can simultaneously authorize multiple operations. This capability has theoretically existed since Ethereum's inception, but ERC-4337 has standardized its implementation and created the alternative mempool infrastructure necessary for production deployment.

The conference on November 18 revealed commitments from mainstream wallets: MetaMask, Argent, and Safe have all deployed smart account infrastructure, with Safe reporting over 100,000 active accounts as of November 2025. Additionally, user experience improvements extend beyond cross-chain operations to include using ERC-20 tokens to pay Gas fees (via a payment manager), social recovery mechanisms, and programmable spending limits.

Fast Finality and 6-Second Slots

In his talk on enhancing Ethereum's service quality as a "confirmation engine" for the entire ecosystem, Barnabé Monnot emphasized two metrics: inclusion time (currently averaging 12 seconds) and final confirmation time (currently 13 minutes). Fast confirmation rules, set to be deployed in Q1 2026, can provide 95% certainty within 1 to 2 blocks (12 to 24 seconds) without waiting for the 13-minute economic finality. This looser security assumption (based on probability rather than economic finality) is sufficient for many use cases: L2s can use confirmed L1 states more quickly (benefiting Base Rollup), cross-chain bridge protocols can achieve faster cross-chain messaging, and centralized exchanges can reduce deposit and withdrawal delays.

Long-term plans include reducing the slot time from 12 seconds to 6 seconds, effectively doubling the network's block rate. Current client performance (processing 60 million Gas blocks in 2.4 to 3.3 seconds) indicates that a 6-second slot is feasible under the premise of raising the Gas limit to 100 million or higher, with continued client optimizations. When combined with fast confirmation rules, a 6-second slot could achieve effective finality within 6 to 12 seconds, making it comparable to centralized payment networks.

DeFi Infrastructure: $300 Trillion Credit Opportunity

Stani Kulechov's Renaissance Finance Thesis

In his talk "The New Architecture of Credit," Stani Kulechov drew historical parallels between financial innovations of the Renaissance and modern DeFi primitives, positioning decentralized credit markets as a $300 trillion opportunity aimed at unlocking global capital flows. In Florence in 1252, the florin became the first widely trusted base currency due to its standardized weight and predictable purity, enabling the expansion of credit across Europe. DeFi replicates this function through stablecoins (USDC, DAI, USDT providing $150 billion in on-chain liquidity at the base layer); Venice's commercial intelligence network collected business information from Mediterranean ports, functionally equivalent to oracle infrastructure (Chainlink providing price feeds and off-chain data verification); commercial networks like the Hanseatic League created liquidity layers connecting local markets, similar to automated market makers (Uniswap, Curve enabling instant token swaps across liquidity pools); and Renaissance commenda contracts allowed passive investors to fund merchant voyages with profit-sharing agreements, serving as precursors to smart contracts that automatically allocate capital based on programmed conditions.

Kulechov's core argument is that the $300 trillion global credit market remains out of reach for DeFi because traditional credit relies on local information (borrower reputation, legal enforceability, collateral assessment), which cannot be directly put on-chain. His proposed solution—the Aave Horizon protocol—tokenizes local credit to enable global DeFi liquidity participation. Local credit analysts use traditional methods (credit history, cash flow analysis, collateral assessment) to evaluate borrowers, then package loans into tokenized tranches for on-chain trading. DeFi liquidity providers purchase these tranches, earning returns from local credit markets, while the protocol handles compliance, collections, and default management.

Kulechov specifically chose Argentina as a case study for the conference. Argentina's credit market exhibits extreme inefficiencies: credit card annual interest rates exceed 100%, and despite valuable real estate, the availability of mortgages remains limited, with capital controls hindering cross-border investment. Furthermore, while institutional investors seek returns in emerging markets, local businesses with strong cash flows struggle to obtain growth capital at reasonable rates. The Aave Horizon pilot project announced during Devconnect aims to bridge this gap by tokenizing accounts receivable from Argentine SMEs and offering them to global DeFi investors at annualized rates of 15% to 25% (attractive to investors and transformative for borrowers accustomed to rates above 100%).

Atomic Settlements and Programmable Compositions

In his talk on institutional adoption, Danny Ryan emphasized the operational improvements brought by blockchain infrastructure through cryptographic settlements (rather than legal enforcement). In traditional finance, stock trades settle on T+1 (one business day), corporate bonds on T+2, and private equity transactions take 90 to 180 days. Each settlement involves multiple intermediaries (transfer agents, custodians, clearinghouses, payment processors), with corporate bond trades requiring about 20 manual steps, and 5% to 10% of trades failing due to reconciliation errors. Ethereum simplifies this to atomic execution: smart contracts receive assets from both parties, either completing the exchange immediately or rolling back the entire transaction. On L2, settlements can be completed within 12 seconds at a cost of under $5, achieving a 99.9% improvement in both time and cost metrics.

More importantly, atomic compositions enable financial products that are impossible in traditional systems. Morpho's presentation showcased cross-collateralized loan products. When an institutional client deposits $100 million in tokenized U.S. Treasuries and immediately borrows $90 million in USDC, the loan terms automatically adjust based on Treasury yields, and if the collateral ratio falls below a safe threshold, programmatic liquidation is automatically executed. The entire process requires no legal contracts, no credit checks, and no settlement delays.

Privacy Infrastructure and Scoped Visibility

The Privacy Summit on November 19 clearly indicated that privacy has become a major barrier to institutional adoption, surpassing regulatory issues in importance. Europe's MiCA provides a clear regulatory framework. The U.S. approval of Bitcoin and Ethereum ETFs also demonstrates that regulators recognize cryptocurrencies as an asset class. However, the development of privacy infrastructure has lagged behind regulatory clarity.

What institutions need, as described by the speakers, is "scoped visibility": different stakeholders see different subsets of data based on their roles and permissions. Fund managers must see complete holdings to make asset allocation decisions. Regulators must be able to verify compliance without accessing strategic trading information. Clients must be able to view their own positions without knowing about others'. The transparency of public chains makes all information visible to everyone, failing to meet these requirements.

In response, the proposed technical solutions layer various cryptographic techniques. Private L2s, such as Aztec, default to encrypting state, with decryption keys distributed according to access policies defined in smart contracts. Zero-knowledge proofs enable selective disclosure: proving to regulators that KYC checks have been passed without revealing identity, or demonstrating that transactions remain within approved limits without exposing actual positions and counterparties. Multi-party computation allows parties to collaborate on analysis without knowing the original input data of others.

BlackRock's BUIDL fund (with assets reaching $500 million as of November 2025) operates on Ethereum but requires permissioned access and off-chain reporting to maintain privacy. The structure of the fund not only demonstrates market demand for blockchain settlement (atomic, programmable, available 24/7) but also exposes the current inadequacies of privacy infrastructure. Multiple presentations indicated that once native privacy features enter production in 2026, they will unleash larger-scale institutional deployments, with tokenized asset volumes potentially reaching $100 billion by 2027.

Conclusion: Infrastructure is in Place, Awaiting Application Takeoff

Devconnect Argentina 2025 indicates that the Ethereum Foundation has decisively shifted towards an application infrastructure-centric approach, rather than treating protocol research itself as the ultimate goal. The five-day technical presentations focused on solving specific problems and provided clear timelines: addressing cross-chain liquidity fragmentation through the EIL deployed by mid-2026; meeting DeFi privacy needs with zkRollup encrypted state going live in 2026; enabling proxy payment infrastructure through x402 and ERC-8004 launched in Q1 2026; satisfying institutional settlement needs through privacy features and throughput expansion; enhancing Blob data availability through PeerDAS in the Fusaka upgrade (December 3, 2025); managing state growth through pragmatic optimizations (client state pruning, witness data compression), while developing quantum-resistant binary tree alternatives for the sustainability era from 2027 to 2030. Each solution targets deployable improvements with measurable success criteria, rather than time-uncertain theoretical advancements.

This approach reflects hard-earned lessons from a decade of operations: early Ethereum's pursuit of ambitious protocol changes consumed years of research and engineering time: sharding technology spanned from 2016 to 2020, after which the focus shifted to a Rollup-centric roadmap; Verkle trees lasted from 2018 to 2025, ultimately abandoned due to quantum computing vulnerabilities; proof of stake extended from 2014 to 2022, only completing with the "Merge" in September 2022. These efforts achieved technical success (or revealed fundamental limitations), but delivery speed lagged behind expectations, while application-layer innovations like the DeFi summer of 2020 and NFT adoption in 2021 demonstrated that significant value could be created on existing infrastructure through careful smart contract design. The current strategy has reversed this priority order.

Protocols like ePBS and BAL serve specific application needs (parallel execution, mitigating MEV) rather than theoretical improvements for their own sake. Privacy features address institutional needs (scope-based visibility for compliance) rather than abstract privacy rights. Cross-chain infrastructure tackles user experience issues (liquidity fragmentation across over 50 L2s) rather than achieving architectural completeness. The willingness to abandon Verkle trees after years of investment exemplifies this pragmatic shift: timely termination of designs with quantum vulnerabilities to avoid losses is preferable to deploying infrastructure that would need replacement within a decade.

Looking ahead, if applications built on this infrastructure can create substantial value for users, this pragmatic foundation will prepare Ethereum for mainstream adoption. Achieving a throughput of 300 million Gas by 2028, deploying privacy-preserving smart contracts in 2026-2027, enabling seamless cross-chain operations through the EIL by mid-2026, and ensuring institutional-grade reliability (100% uptime since September 2022)—the combination of these capabilities creates an infrastructure capable of supporting trillions of dollars in economic activity. Whether this potential is realized depends on factors beyond the control of the Ethereum Foundation: regulatory evolution in major jurisdictions, competitive dynamics with other public chains offering different trade-offs, and most critically, whether applications built on this infrastructure address the real problems faced by users rather than the interesting problems perceived by engineers. However, from a purely infrastructural perspective, the decade-long evolution from idealism to pragmatism has created a platform ready for serious economic activity. It has traded philosophical purity for deployable solutions, abstract decentralization for measurable security attributes, and revolutionary ambition for incremental compound growth.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。