The turning point of the 10-year debate: Can Ethereum truly solve the "trilemma"?
Written by: imToken
The word “trilemma” makes everyone's ears callouse, right?
In the first decade of Ethereum's birth, the “trilemma” was like a physical law hanging over every developer's head - you can choose between decentralization, security and scalability, but you can never have all three. However, looking back at the beginning of 2026, we will see that it seems to be gradually becoming a “design threshold” that can be crossed through technological evolution, as Vitalik Buterin pointed out on January 8 as a disruptive view:
Compared to reducing latency, increasing bandwidth is more secure and reliable, and with PeerDAS and ZKPs, Ethereum's scalability can increase thousands of conflicts, which is not twice as high as decentralization.
- Decentralization means low node threshold, widespread participation, and no need to trust a single subject;
- Security means that the system remains consistent against evil, censorship, and attacks;
- Scalability, which means high throughput, low latency, and a good user experience;
The problem is that these three often constrain each other under traditional architectures, such as increasing throughput, which usually means raising hardware thresholds or introducing centralized coordination. Reducing the burden on nodes may weaken security assumptions; Insisting on extreme decentralization inevitably sacrifices performance and experience. It can be said that in the past 5-10 years, from the early EOS to the later Polkadot and Cosmos, to the ultimate performance seekers Solana, Sui, Aptos, etc., different public chains have given different answers. But the common denominator isAlmost all expansion plans can only meet two of them at the same time, and the third must be sacrificed. Or to put it another way, almost all solutions are repeatedly saw-saw under the logic of “single blockchain” - if you want to run fast, you have to have strong nodes; If you want to have many nodes, you have to run slowly, which seems to have become a dead proposition. If we put aside the debate about the advantages and disadvantages of monolithic module blockchains for a moment and carefully review the development path of Ethereum's comprehensive shift from a “monolithic chain” to a “rollup-centric” multi-layer architecture in 2020, as well as the recent maturity of supporting technologies such as ZK (zero-knowledge proofs), we will find: The underlying logic of the “trilemma” has been slowly reconstructed in the past five years in the era of Ethereum's modularization. Objectively speaking, Ethereum decouples the original limitations one by one through a series of engineering practices, at least in terms of engineering paths, this issue is no longer just a philosophical discussion. 2. The engineering solution idea of “divide and conquer” Next, we will break down these engineering details, specifically how Ethereum has been advancing in parallel through multiple technical lines to resolve this triangular limitation in the five-year demonstration from 2020 to 2025. The first is to “decouple” from data availability through PeerDAS, liberating the inherent ceiling of scalability. As we all know, in the trilemma, data availability is often the first shackle that determines scalability, because traditional blockchains require each full node to download and verify all data, ensuring security while also limiting the scalability limit. The direction given by Ethereum is not to make nodes stronger, but to change the way nodes verify data, which uses PeerDAS (Peer Data Availability Sampling) as the core solution:
It no longer requires each node to download all block data, but verifies whether the data is available through probabilistic sampling - block data is split and encoded, and nodes only need to randomly sample part of the data.
Vitalik emphasized that PeerDAS is no longer a concept in the roadmap, but a real-world deployed system component, which also means that Ethereum has taken a substantial step in “scalability × decentralization”. The second is zkEVM, which attempts to solve the problem of “does each node have to repeatedly perform all calculations” through a zero-knowledge proof-driven verification layer. Its core idea is to equip the Ethereum mainnet with the ability to generate and verify ZK proofs. In other words, after each block is executed, it can output a verifiable mathematical proof, allowing other nodes to confirm the correctness of the result without repeated calculations.
- Faster Verification: Nodes do not need to repeat transactions, only need to verify zkProof to confirm block validity;
- Lighter burden: Effectively reduces the pressure on full-node computing and storage, making it easier for light nodes and cross-chain validators to participate.
- Stronger security: Compared to OP routes, ZK's state proofs are confirmed in real time on-chain, with higher tamper-resistant capabilities and clearer security boundaries.
Not long ago, the Ethereum Foundation (EF) officially released the L1 zkEVM real-time proof standard, marking the first time that the ZK route has been officially written into the mainnet-level technical plan. According to EF's announced technical route, the block proof latency target is controlled within 10 seconds, the size of a single zk proof is less than 300 KB, and the 128-bit security level is adopted, avoiding trusted setup, and planning to allow home devices to participate in proof generation to lower the threshold for decentralization. Finally, in addition to the above two, there are also Ethereum roadmaps based on the pre-2030 (such as The Surge, The Verge, etc.), which revolve around increasing throughput, refactoring state models, raising gas limits, and improving the execution layer. These are trial-and-error and accumulation paths that cross traditional triangular constraints, and are more like a long-term thread, committed to achieving higher blob throughput, clearer rollup division of labor, and more stable execution and settlement rhythms, laying the foundation for future multi-chain collaboration and interoperability. Importantly, these are not isolated upgrades, but are clearly designed as modules that superimpose and complement each other, which also reflects Ethereum's “engineering attitude” to the trilemma: not to find a magic solution to win like a single blockchain, but to redistribute costs and risks through multi-layer architectural adjustments. 3. Vision 2030: The final shape of Ethereum Even so, we need to exercise restraint. Because factors such as “decentralization” are not static technical indicators, but the result of long-term evolution. Ethereum is actually using engineering practice step by step to explore the constraints of the trilemma——With the change in verification methods (from recalculation to sampling), data structure (from state expansion to state expiration), and execution model (from monolithic to modular), the original trade-off relationship is shifting, and we are infinitely approaching the end point of “both wanting, wanting, and wanting”. In recent discussions, Vitalik also gave a relatively clear time frame:
- 2026: With some execution layer/construction mechanism improvements, after the introduction of ePBS and other directions, the gas limit that does not rely on zkEVM can be increased first, while creating conditions for “more widespread operation of zkEVM nodes”.
- 2026–2028: Adjustments around gas pricing, state structure, and execution load organization to enable the system to operate safely under higher loads;
- 2027–2030: As zkEVM gradually becomes an important way to validate blocks, the gas cap may be further increased, and the long-term ideal goal is to more decentralized block building;
- Minimalist L1:**L1 becomes a solid, neutral layer that is only responsible for providing data availability and settlement proofs, and it no longer handles complex application logic.**Thus maintaining an extremely high level of security;
- Thriving L2s and Interoperability: Through EIL (Interoperability Layer) and quick confirmation rules, fragmented L2s are stitched into a single whole,Users cannot perceive the existence of the chain, only a TPS of 100,000 levels;
- Extremely low validation threshold: Due to the maturity of state processing and light client technology,Even mobile phones can participate in verification, which ensures that the cornerstone of decentralization is as solid as Mount Tai;
Interestingly, at the time of writing, Vitalik once again emphasized an important testing standard - “The Walkaway Test”, reiterating that Ethereum must have the ability to operate autonomously, even if all server providers disappear or are attacked, DApps can still operate and user assets remain secure. This sentence is actuallyPull the evaluation scale of this “endgame pattern” back from speed/experience back to what Ethereum cares about most - that is, whether the system is still trustworthy and still does not rely on a single point in the worst case. Written at the end People always have to look at problems with a development perspective, especially in the ever-changing industry of Web3/Crypto. The author also believes that many years later, when people think back to the fierce debate about the trilemma of 2020-2025, they may feel that it is like before the invention of the automobile, people were seriously discussing “how horse-drawn carriages can balance speed, safety, and load at the same time”. The answer given by Ethereum is not to do painful multiple-choice questions between the three vertices, but to build a digital infrastructure that belongs to everyone, is extremely secure, and can carry the financial activities of all mankind through PeerDAS, ZK proofs, and exquisite economic game design. Objectively speaking, every step in this direction is stepping on the end of the past of the “trilemma”.