zk-evm

Ethereum co-founder Vitalik Buterin posted on the X platform that Ethereum's scalability is divided into short-term and long-term parts.In the short term, the Glamsterdam upgrade will introduce block-level access lists for parallel validation, ePBS will allow a larger proportion of time slots for block validation, and gas repricing will ensure that operational costs align with actual execution time.The multi-dimensional gas mechanism will be implemented in phases, starting with Glamsterdam, where the "state creation" cost will be separated from the "execution and calldata" cost, with state creation gas not counted towards the approximately 16 million transaction gas limit. The EVM level will introduce a "reservoir" dimension mechanism, which will prioritize the consumption of dedicated dimension gas by default, and when insufficient, will draw from the reservoir. This will eventually transition to multi-dimensional pricing, where different dimensions may have different floating gas prices.Long-term scalability includes ZK-EVM and blob components. In terms of blobs, there are plans to continuously iterate PeerDAS, aiming to achieve approximately 8MB of data processing capacity per second, with future Ethereum block data directly entering blobs.The ZK-EVM aspect will be implemented in phases: by 2026, there will be a validator client supporting ZK-EVM, allowing about 5% of the network to rely on it; by 2027, this will expand to a larger proportion of a few nodes while advancing formal verification; once conditions are mature, it will transition to a five-out-of-three mandatory proof mechanism, ultimately continuously enhancing the security and formal verification level of ZK-EVM, and involving changes to VMs such as RISC-V.

Vitalik Buterin posted on the X platform that he is trying to push forward a more ambitious plan: to build an "Ethereum that follows cyberpunk principles and is not ugly" as an extension layer on top of the existing system, while achieving tight integration and high interoperability with the current framework as much as possible.The proposal will develop gradually, while ensuring that Ethereum itself acquires the necessary system-level characteristics of cyberpunk and simplicity, including censorship resistance, friendliness to zero-knowledge proofs, and attributes of the consensus layer.Vitalik Buterin further stated that within the next 5 years (or even faster in the case of accelerated AI coding and formal verification), an open path may be established, allowing the existing system to migrate to smart contracts written in the new system language when needed.Ethereum has previously completed significant upgrades during its operation, akin to "changing jet engines in mid-air"—such as the "Merge"—and still has the capacity for about four more system-level transformations of similar scale in the future, including state tree optimization, streamlined consensus, ZK-EVM verification, and changes to the virtual machine layer.

Vitalik Buterin stated that he is "significantly more inclined to support native rollups" compared to the past. Vitalik recalled that a major reason for opposing native rollups previously was that their precompiled solutions had to choose between ZK mode or Optimistic mode, and at that time, ZK-EVM was not mature. L2 often chose the latter, which involved "fast withdrawals but self-proving risks" versus "relying on Ethereum's security but needing to wait 2-7 days for withdrawals," thereby weakening Ethereum's composability and promoting the prevalence of multi-signature bridges and other solutions.Vitalik pointed out that the situation is changing: Ethereum's timeline for fully adopting ZK at the L1 level is gradually aligning with the realistic progress of introducing native rollup precompiles, and the aforementioned core obstacles are expected to be eliminated. He also mentioned that the community is increasingly viewing "synchronous composability" as one of the core values of L2 and is exploring the combination of rollup-based solutions with low-latency pre-confirmation mechanisms.In addition, Vitalik emphasized that the design of native rollup precompiles should not be rushed in its implementation. He expressed a hope that in the future, there would be a property such that if developers build a rollup that is "EVM plus a small amount of extended functionality," they could directly reuse the EVM part of the native rollup precompiles and only introduce a custom proof system for the new features, connecting the two in a standardized way.

Vitalik Buterin posted on social media that 2026 will be a key year for Ethereum to regain lost ground in self-sovereignty and trustlessness. The plans include simplifying node operation through ZK-EVM and BAL technology, launching Helios to verify RPC data, implementing ORAM and PIR technology to protect user privacy, developing social recovery wallets and time-lock features to enhance fund security, and improving on-chain UI and IPFS applications. Buterin emphasized that Ethereum will correct the compromises made in the past decade regarding node operation, application decentralization, and data privacy, refocusing on core values. Although this will be a long process, it will make the Ethereum ecosystem stronger.

Vitalik Buterin stated that Ethereum is close to solving the long-standing problem of the "blockchain trilemma" at the practical operational level. He pointed out that two key upgrades, PeerDAS and ZK-EVM, are making Ethereum a "brand new, more powerful decentralized network."Vitalik mentioned that PeerDAS has been launched on the mainnet in 2025, while ZK-EVM, although still in the phase of enhancing security, has reached production-level performance and is expected to begin small-scale usage in the network in 2026. He further outlined that in the coming years, Ethereum will gradually achieve a balance between decentralization, security, and high throughput by increasing the gas limit, adjusting the state structure, and introducing more ZK-EVM-based verification methods. He emphasized that this is not just a theoretical concept, but a long-term engineering achievement built on operational code.Vitalik also reflected that Ethereum has invested nearly 10 years in addressing data availability and scalability issues, and this vision is now gradually coming to fruition.

Vitalik Buterin stated that the current ZK-EVM has entered the alpha stage, demonstrating production-level performance, with the remaining major work focused on security; meanwhile, PeerDAS has officially launched on the Ethereum mainnet. Vitalik pointed out that this is not a gradual optimization, but rather a transformation of Ethereum into a new type of P2P network that simultaneously possesses decentralization, consensus mechanisms, and high bandwidth.He recalled that BitTorrent has high bandwidth but lacks consensus, while Bitcoin has consensus but is bandwidth-limited due to full replication; with the introduction of PeerDAS (data availability sampling) and ZK-EVM, Ethereum is expected to achieve decentralization, consensus, and high throughput simultaneously, solving the trilemma through "real code running on the mainnet." Among these, PeerDAS is already running on the mainnet, and ZK-EVM has reached production-level performance.Vitalik expects that starting in 2026, with the advancement of mechanisms like BAL and ePBS, the gas limit will gradually increase, and there will be practical opportunities to run ZK-EVM nodes; from 2026 to 2028, there will be gas repricing, state structure adjustments, and execution load entering blobs; by 2027 to 2030, ZK-EVM is expected to become the primary method for validating blocks on the network, further promoting an increase in the gas limit. Additionally, he mentioned the importance of distributed block construction, with the long-term goal of avoiding the complete construction of blocks in any single location and decentralizing block construction power through on-protocol or off-protocol means to reduce the risk of centralized intervention and enhance regional fairness.

Vitalik Buterin tweeted that the modexp precompile is extremely unfriendly to ZK-EVM, with the worst-case scenario reaching an average of 50 times the computational overhead per block. He called for an EIP to replace it with equivalent EVM code, even if it requires higher gas. He admitted that as the designer of modexp, he feels "ashamed" and pointed out that its usage is very low and increases the risk of consensus failure, stating that the efficiency of the entire Ethereum ecosystem should not be sacrificed for a very small number of needs.

ChainCatcher news, Vitalik Buterin posted on the X platform expressing his high appreciation for Sandeep Nailwal and Polygon's key role in the Ethereum ecosystem, particularly in the early investment in ZK-EVM, supporting highly scalable applications (such as Polymarket), and promoting pandemic research projects. He also revealed that Sandeep had returned a donation of $190 million worth of SHIB, providing crucial funding support for the Balvi project. Vitalik encouraged Polygon to adopt off-the-shelf ZK technology in the future to achieve higher security guarantees.

ChainCatcher news, regarding the multi-GPU zero-knowledge virtual machine (zkVM) Pico Prism released by Brevis, Vitalik commented, "I am glad to see Brevis's Pico Prism officially entering the ZK-EVM verification field. The verification speed and diversity of ZK-EVM have taken an important step forward."In previous news, Pico Prism has achieved real-time proof of Ethereum on consumer-grade hardware: using 64 RTX 5090 graphics cards, it completed 99.6% of Ethereum L1 block proofs within 12 seconds, with 96.8% of the block proof times falling below the 10-second standard set by the Ethereum Foundation. In a test on September 1, under Ethereum's current gas limit of 45M, Pico Prism had an average proof time of only 6.9 seconds.

ChainCatcher news, Vitalik Buterin published a new article titled "Exploring circle STARKs," which points out that Starkware can prove 620,000 Poseidon2 hashes per second on an M3 chip laptop. This means that if we are willing to trust Poseidon2 as a hash function, one of the most challenging parts of creating an efficient ZK-EVM has actually been solved.It states that compared to conventional STARKs, circle STARK does not introduce too much additional complexity for developers. The mathematical principles behind the "polynomials" operated by circle FRI are quite counterintuitive and take some time to understand and grasp. However, this complexity is precisely hidden, and developers remain unaware of it. The complexity of Circle's mathematical principles is encapsulated rather than systemic.Vitalik believes that by combining Mersenne31, BabyBear, and binary field technologies like Binius, we are approaching the efficiency limits of STARKs' "base layer"; it is expected that the frontier of STARK optimization will shift towards the most efficient arithmetic of primitives such as hash functions and signatures (and optimizing these primitives themselves for this purpose), creating recursive structures for more parallelization, arithmetic for virtual machines to improve developer experience, and other higher-level tasks.

ChainCatcher news, Ethereum co-founder Vitalik Buterin published an in-depth discussion on the concept of "ZK-EVM" (Zero-Knowledge Ethereum Virtual Machine) and its possible implementations.The article points out that current Layer-2 EVM protocols (such as Optimistic Rollups and ZK Rollups) rely on the verification mechanism of the EVM, but this also means they must trust a large codebase. Once there are vulnerabilities in the codebase, these virtual machines may face the risk of being attacked. Furthermore, even the ZK-EVM, which aims to maintain complete equivalence with L1 EVM, requires some form of governance mechanism to replicate changes from L1 EVM into its own EVM implementation.The ZK-EVM concept proposed by Buterin aims to reduce the redundant implementation of Ethereum protocol functionalities by Layer-2 projects and improve their efficiency in validating Layer-1 Ethereum blocks. He also envisions that future light clients will become more powerful and may even utilize ZK-SNARKs (zero-knowledge proofs) to fully verify L1 EVM execution, at which point the Ethereum network will essentially have built-in ZK-EVM functionality.The article also discusses different versions of implementing ZK-EVM, including their design challenges, trade-offs, and why certain directions may not be adopted. It emphasizes that when implementing protocol functionalities, one should weigh the benefits against the advantages of keeping the underlying protocol simple. Regarding the key attributes of ZK-EVM, Buterin highlighted its fundamental functionality, compatibility with Ethereum's multi-client philosophy, data availability requirements, auditability, and upgradability. Additionally, he mentioned support for "almost-EVM," which allows L2's VM to use the protocol's ZK-EVM while having only minor differences from EVM, providing flexibility for partial customization of EVM.

ChainCatcher news, the official of the blockchain game Abyss World stated on social media that the Abyss World team has decided to build the game based on Polygon zk-EVM.It is reported that this will make the execution of smart contracts in Abyss World more efficient and capable of handling more user interactions. The team hopes that the collaboration with PolygonGaming will unleash the full potential of creativity in Web3 gaming. (Source link)