In a consumer-grade hardware environment, 64 GPUs completed 99.6% of the L1 block proof in 12 seconds—Brevis's multi-GPU zero-knowledge virtual machine (zkVM) Pico Prism has achieved a significant performance breakthrough.
This milestone has also attracted significant attention from the Ethereum core community. Ethereum founder Vitalik Buterin publicly commented, "I am glad to see Brevis's Pico Prism officially entering the ZK-EVM verification field. This is an important step for ZK-EVM in terms of proof speed and diversity."
"The data speaks for itself." Brevis co-founder and CEO Michael stated, "We have built infrastructure capable of handling Ethereum's real-time block output, all using consumer-grade hardware. This performance is the best response to Ethereum's goal of decentralization."
Brevis co-founder and CEO Michael—PhD in Computer Science from UIUC, expert in distributed systems and high-performance networks—has successfully launched multiple startups in Silicon Valley. His research has been adopted by many tech companies, including Google. In 2018, he founded Celer Network (with a peak valuation of over $2 billion). Now, he leads the Brevis team to combine zero-knowledge proofs with verifiable computing, hoping to create infrastructure that can truly support the large-scale computing needs of the Ethereum ecosystem.
To better understand how Brevis achieved this breakthrough and what it means for the Ethereum ecosystem, Rhythm BlockBeats conducted an in-depth conversation with Brevis co-founder and CEO Michael.
We explored three core questions:
- Brevis's mission and positioning
- The practical value verifiable computing can bring to applications
- Brevis's next steps in aligning deeply with the Ethereum ecosystem
This is not just a conversation about technology, but also a forward-looking reflection on the future of computational trust.
What is Brevis: From Vision to Positioning
Rhythm BlockBeats: Could you please introduce Brevis in the simplest terms? What kind of project is it, and what problems does it aim to solve?
Michael: We position Brevis as the Infinite Computing Layer, serving Web3 and covering a broader range of application scenarios. The core goal is to provide nearly infinite computing power for on-chain applications while maintaining decentralization and security, addressing the structural contradiction between high performance/complex functionality and long-term trust minimization. Although L2 scaling has significantly improved throughput in recent years, complex computations (such as large-scale historical data processing, cross-chain aggregation, algorithm/AI inference, etc.) are still limited, making it difficult to implement many advanced features.
Brevis's solution is off-chain computation and on-chain verification. Heavy computations are migrated off-chain, generating zero-knowledge proofs (ZK Proof); on-chain contracts only need to verify the correctness of the proof at a very low cost, without needing to re-execute the entire computation. Thus, the contract effectively gains a cryptographically secured plugin, significantly expanding the available computing power and functional boundaries without sacrificing decentralization and security. Smart contracts are not very smart right now, but with Brevis, they can truly become intelligent.
Rhythm BlockBeats: Brevis is referred to as the infinite computing layer of Web3. Where does this concept come from? How is it fundamentally different from traditional on-chain computation or scaling?
Michael: We call Brevis the infinite computing layer of Web3 because its goal is to break the computational limits of blockchain. Traditional scaling solutions (like Layer 2 Rollup) mainly enhance transaction throughput, increasing from dozens of TPS to hundreds or thousands of TPS, but there is still a ceiling and they do not address whether any complex computation can be completed on-chain: on-chain typically can only handle relatively simple transactions and contract operations. Once the computational complexity increases, it becomes difficult to model and bear on-chain, such as large-scale data processing, AI inference, or complex algorithms.
Brevis's starting point is different: we are not just making established types of transactions run faster, but enabling the blockchain to accept and carry any computational results while maintaining security and trustlessness. This corresponds to a new paradigm of verifiable computing— as long as off-chain computation can produce a zero-knowledge mathematical proof, the on-chain can confirm its correctness and security at a very low cost.
For developers, Brevis is more like a cloud of verifiable computing: scaling computing power on demand like using cloud computing, completing complex computations off-chain, and then submitting the proof for on-chain verification, while retaining a security and trust model consistent with native on-chain contracts. Therefore, we define Brevis as an infinite computing layer—achieving nearly infinite computational space in terms of performance and usability, while maintaining consistency in trust and security with the blockchain itself.
Rhythm BlockBeats: The verifiable computing you just mentioned sounds abstract. Can you explain its role in Brevis in a more intuitive way?
Michael: The fundamental limitation of blockchain lies in its limited computing power while ensuring security and trustlessness. Public chains like Ethereum belong to consensus systems, where a large number of nodes in the network compute the same transaction or contract execution separately, reaching consensus before producing a new block. This mechanism is secure but inefficient, similar to having the entire class re-solve the same problem from scratch to confirm the answer.
The idea of verifiable computing is to decouple computation from verification, reducing redundant overhead. For example, in the case of the "chickens and rabbits" problem: solving it requires setting up equations and deriving results, while verifying whether a given answer is correct only requires substituting values back into the equation, which is much cheaper than re-solving it. Computation itself and verifying the results of computation are inherently two different problems in computer theory, with the latter usually being lighter. Zero-knowledge proofs extend this to any computation, whether it's a simple operation like 1+1=2 or inference from a large model, allowing for a concise proof to be generated after computation, enabling others to quickly confirm correctness at a very low cost without revealing computation details.
In Brevis, heavy computations are executed off-chain and generate ZK proofs, while on-chain only verifies the proofs without needing to re-execute the entire computation. This is a core value of verifiable computing, as it significantly saves the overhead of repeatedly computing the same target.
Why Do This: Technical Architecture and Core Ideas
Rhythm BlockBeats: Technically, Brevis consists of two core modules: the ZK Data Coprocessor (hereinafter referred to as zkCoprocessor) and the Pico zkVM. What is the relationship between the two? What problems do they each solve?
Michael: Brevis can be understood as a layered structure: the underlying Pico zkVM is a general-purpose computing engine (Virtual Machine) that can execute any computation and generate zk Proof; one of its important features is high modularity, allowing it to connect multiple coprocessors. The zkCoprocessor is a coprocessor tailored for blockchain data scenarios, which you can think of as a plugin or add-on. In its current form, the zkCoprocessor acts more like a memory system for the zkVM, enabling smart contracts to see and understand historical events on-chain (such as user transactions, balances, holdings, etc.).
More specifically, the zkVM addresses the general problem of making any computation verifiable; while the zkCoprocessor is an application-layer coprocessor customized for blockchain applications. Since smart contracts inherently "live in the present," they can only access the limited context of the current block and cannot directly read or compute historical or cross-chain states over long time windows; the zkCoprocessor gives contracts "eyes" through zero-knowledge proofs, retrieving and aggregating relevant historical data off-chain and generating proofs that these data are real and derived from on-chain states, thus enabling contracts to possess memory and memory-based intelligence while maintaining trust minimization.
Rhythm BlockBeats: The zkCoprocessor allows smart contracts to "see the past," which sounds groundbreaking. Can you elaborate on how it achieves this?
Michael: We can use PancakeSwap's dynamic fee discount based on historical trading volume as an example:
First, the zkCoprocessor reads the user's historical transaction records from the blockchain, aggregates the trading volume according to rules (such as the last 30 days), and generates a zero-knowledge proof for it;
Then, the Pico zkVM, upon receiving the verified records, performs further aggregation and computation (for example, merging the trading volumes of different trading pairs into a single metric) and generates a proof for the computation process;
Finally, the results + proof are submitted on-chain, and the contract verifies the correctness of both proofs at once, confirming whether the address has reached the VIP threshold and automatically applying the fee rate discount for the next settlement period based on that.
In this way, the contract does not need to repeatedly read and compute massive historical data on-chain, gaining the ability to see the past while maintaining trustlessness and cost control.
Rhythm BlockBeats: You developed the Pico zkVM yourself instead of adopting solutions from other zkVM projects. What considerations were behind this? How does Brevis differ from other zk projects?
Michael: This is a very critical question. Indeed, there are already many zkVM projects on the market, so why did Brevis choose to build its own Pico zkVM from scratch? The core reason is quite simple—we believe that most existing zkVMs are still in the experimental stage and are not truly designed for large-scale, real-world applications. In terms of overall performance cost and scalability, they are still far from being able to support large-scale commercial use. In other words, they are more like proof of concepts rather than production-grade systems capable of supporting millions of calls.
Brevis faced very realistic demand scenarios from the beginning. Leading DeFi protocols like PancakeSwap, Euler, Linea, and Usual need to generate millions of zk proofs every day. If the underlying zkVM's performance does not meet the standards, the entire system cannot go live. This forced us to rebuild a zk computing engine that is truly born for production environments—this is the Pico zkVM.
Pico has three notable features. First is extreme performance. Pico is currently the world's most powerful zkVM, and our latest release, Pico Prism, can achieve real-time proof of 99.6% of Ethereum mainnet blocks on 64 5090 GPUs (completed within 12 seconds), with 96.8% of blocks even completed in under 10 seconds. This means a performance improvement of 3–4 times and a reduction in hardware costs of about 50%. In other words, if Ethereum were to run on Pico today, its overall verification efficiency could improve by an order of magnitude, representing a true breakthrough in real-time proof.
Second is a unique modular architecture. Pico is currently the only zkVM that supports external coprocessors; other zkVMs typically have a closed structure and can only perform general computations, while Pico can mount different modules based on various application scenarios. For example, when historical data access, cross-verification of on-chain records, or running complex financial logic is required, Pico's coprocessors can be plugged in to accelerate specific tasks. This gives Pico both generality and specialized high performance, allowing it to flexibly cover a variety of applications from Web3 DeFi to Web2 AI computing.
Finally, it is developer-friendly. We do not want developers to have to be experts in cryptography or ZK theory to use Pico. Developers only need to know how to write Rust to directly call and write ZK applications just like writing ordinary programs. This greatly lowers the barrier to entry and completely hides the complexity of ZK at the underlying level. So overall, we are not an experimental zkVM, but a ZK computing engine designed for the real world and real use cases.
Rhythm BlockBeats: Recently, you announced that Pico Prism achieved 99.6% real-time proof on consumer-grade hardware. Can you talk about the technical breakthroughs behind this achievement and what it means for the performance limits of zkVM?
Michael: Let me first clarify the importance of this matter. Performance improvement is not merely a technical optimization; it is related to the future scalability and development of the Ethereum ecosystem. The current architecture of Ethereum requires each node to repeatedly execute the same computation, which is secure and reliable, but scalability is nearing its limit. The next step requires a new paradigm: a single node completes a large amount of computation and generates ZK Proof, while other nodes are only responsible for verification. By adopting this single-point computation and multi-point verification approach, we can increase throughput by an order of magnitude while maintaining decentralization and security, potentially reaching or exceeding Solana's level, while maintaining the necessary scale of decentralized nodes.
The Ethereum Foundation proposed a two-year goal in July of this year: to achieve real-time proof of 99% of blocks using consumer-grade hardware costing less than $100,000. Once achieved, it would provide nearly infinite scalability by leveraging the computing power of ordinary hardware. However, this had previously been seen as a theoretical possibility: most zkVM solutions still have significant gaps in coverage, cost, and speed, with many even struggling to achieve 90% real-time coverage.
Pico Prism is the first system to truly cross this performance line. In our tests, we achieved 99.6% of Ethereum blocks being proven within 12 seconds using only 64 5090 GPUs (costing about $120,000), with 96.8% of blocks completed in under 10 seconds and an average proof time of just 6.9 seconds.
Compared to other protocols, Pico's performance improved by 70%, and costs were reduced by 50%. In other words, it is not only faster but also cheaper, which means we are very close to the goals set by the Ethereum Foundation.
Rhythm BlockBeats: In terms of functionality and goals, does Brevis resemble more of a ZK acceleration layer for Ethereum or a verifiable computing cloud aimed at multiple chains?
Michael: In terms of functionality and goals, these two positions are not in conflict. In the short term, Brevis resembles a ZK acceleration layer for Ethereum and its second layer. We align with the Ethereum Foundation's direction: achieving a 10x to 100x scaling of the mainnet through zero-knowledge proofs; at the same time, the real-time proof capability will significantly enhance the interactivity of second-layer Rollups, making cross-chain speeds faster, efficiency higher, and costs lower, while promoting better unification of state and liquidity.
From a longer-term and more macro perspective, Brevis's positioning is not limited to Ethereum. Our architecture naturally supports multiple chains, and we are currently collaborating with ecosystem partners like BNB Chain (such as PancakeSwap) and advancing projects like perpetual contracts on non-Ethereum chains; we also maintain alignment with second-layer/multi-chain ecosystems like Arbitrum, Base, and Avalanche. Overall, we hope to build Brevis into a verifiable computing cloud for all blockchain applications.
Our goal in the next 3–5 years is to enable smart contracts deployed on any chain to easily use this computing service; looking further ahead, the vast majority of computations for blockchain applications will occur off-chain, with ZK methods ensuring trustworthiness, and Brevis will serve as the unified computing trust layer for the entire decentralized system.
How to Implement: Real Applications and Cases
Rhythm BlockBeats: Many DeFi applications, such as PancakeSwap, Euler, or Linea, are already using Brevis's technology. Can you provide a specific example of how it works in these scenarios?
Michael: Yes, our current application scenarios are mainly divided into several categories, with DeFi being the most typical. The aforementioned PancakeSwap is a standard case; we do not want the user experience of DEX to remain at a one-size-fits-all stage—whether you are an ordinary trader or a high-frequency trader, the interface and fees are completely the same. We hope DEX can gradually provide differentiated experiences for different types of users, similar to CEX. At the same time, we are exploring more innovative applications. For example, collaborating with SocialFi projects to allow users to prove their influence and real holdings without exposing their main wallet.
We are also advancing some scenarios that operate almost entirely off-chain. For instance, in perpetual trading systems like Hyperliquid, users' positions, leverage, and other information are often public and can easily be attacked by other traders. We hope to reconstruct such systems using ZK technology, allowing all orders and position records to be encrypted and hidden while still ensuring the correctness of matching and the security of the system through ZK Proof. This can provide a smooth experience close to centralized trading platforms while maintaining blockchain-level security and privacy.
Rhythm BlockBeats: On the product level, Incentra is an important application you launched, which currently manages over $300 million in incentive distribution. How does this system work?
Michael: The starting point of Incentra is simple: incentive distribution is at the core of every ecosystem's growth, but for a long time, the distribution mechanism has often faced issues of security, compliance, and transparency.
The core mechanism of Incentra is that users generate ZK proofs based on their real behavior on-chain or within the protocol, completing qualification verification and claiming directly on-chain without relying on centralized distribution. Compared to traditional methods, this model has advantages in three aspects. First, security: funds are settled within the contract according to preset rules, reducing the risks of centralized custody and operation; second, compliance and auditability: incentive rules and qualification proofs are traceable on-chain, avoiding compliance pressure from payments to unknown entities; third, fairness and transparency: any claiming action can be independently verified by external parties for its corresponding real contribution and calculation path, eliminating opaque distribution. Currently, this system has been implemented in the growth and incentive scenarios of projects like BNB Chain, MetaMask, OpenEden, and Usual. Whether in stablecoin ecosystems or on-chain incentive systems, this solution can achieve secure, compliant, and transparent incentive distribution.
Rhythm BlockBeats: Compared to many zk projects that are still in the experimental stage, Brevis has already achieved large-scale implementation. Why do you think you were able to do this faster?
Michael: I believe the core reason is that from day one, we did not see ourselves as a research-focused ZK laboratory, but as an infrastructure company aimed at real applications. Our technical approach is not to build the engine first and then find a use for it, but rather the reverse: starting from application needs and working backward to design the underlying architecture.
This fundamentally distinguishes Brevis. Many ZK projects first develop a seemingly perfect zkVM behind closed doors and then think about where it can be applied; whereas we start from the real pain points of our partner projects, first understanding their actual needs, and then gradually designing a modular zkVM architecture, further creating a pluggable zkCoprocessor.
In other words, Brevis's technological evolution is not planned out of thin air, but is pushed forward step by step by real users and partners. This problem-oriented R&D model allows us to continuously iterate quickly, with each improvement directly corresponding to a verifiable use case. This also gives our system production-level stability and scalability from the very beginning, rather than writing code for papers.
Another more straightforward reason is that our team understands both cryptography and large-scale systems engineering. We have never stayed at the theoretical level to do derivations, but have genuinely honed performance and verified stability in production environments. So far, Brevis has generated over 100 million zero-knowledge proofs on the mainnet, serving about 190,000 users and managing over $300 million in real incentive funds for partner protocols. These are real applications that are live— including projects like Linea, Euler, Usual, and OpenEden—not demos or testnets.
Where to Go Next: Open Scenarios and Future Vision
Rhythm BlockBeats: Besides DeFi, in what other areas do you think the off-chain computation and on-chain verification model can have an impact, such as AI, data markets, or gaming?
Michael: Beyond DeFi, this model will have a lasting impact in areas such as verifiable AI, bridging Web2 to Web3 data and identity, privacy data markets, and gaming and social applications. First, in the field of verifiable AI, existing models are often black boxes. Through zero-knowledge proofs, verifiable mathematical evidence can be generated for the inference process, proving that a specific model produces the corresponding output for a given input, while ensuring that the model and process have not been replaced or tampered with. Second, in bridging scenarios from Web2 to Web3, users can bring qualifications or behaviors from centralized platforms into on-chain applications in the form of proofs, such as mapping trading activity on centralized trading platforms to fee discounts on decentralized trading platforms without exposing identity or underlying account details. Finally, in gaming and social contexts, users can prove their historical achievements, asset holdings, or completion of key actions to gain access, matching qualifications, or incentives without disclosing wallet or personal sensitive information.
In the long run, this system will redefine the trust boundaries of the digital world. We hope to reconstruct the way human society understands trustworthy computing through Web3 and zero-knowledge proofs, allowing every computation to be verified and trusted. What Brevis aims to do is to become the underlying layer of all this—a new layer of computational trust.
Rhythm BlockBeats: We saw that Brevis launched the Brevis Proving Grounds event on October 13. Can you share what the core content of this event is and what users can truly participate in and experience?
Michael: In fact, we have spent a lot of effort recently on validating applications and document approvals for implementation. We are now deeply collaborating with over twenty partners, and each project is a real, live application that users can directly use, with excellent data performance and user retention.
However, we found that in the blockchain world, many things are actually driven by user understanding, which in turn pushes development. In other words, when users truly understand the value of a new technology, their feedback and demands will drive developers to implement more features. Because of this, we want to truly bring users in, allowing them to not just be bystanders but to personally experience and validate the technology. This is the core purpose of the Brevis Proving Grounds event we launched.
The entire event is roughly divided into two phases. The first phase is the user education phase, where we hope to help more people understand what our partners are doing and what specific problems zero-knowledge proofs (ZK) can solve in these projects. This phase is more about popularization and understanding, making it clear the logic and significance behind this technology.
The second phase is the real experience phase, where users can personally use the applications we have already implemented and truly feel the changes brought about by integrating with Brevis. Unlike many projects, where activities usually just require users to click a mouse, complete tasks, and earn rewards, because they often lack truly experiential, foundational innovative features, we are different. Our partner projects are real, valuable applications, and users are not playing a simulated experience; they are genuinely using products that are deeply integrated with Brevis technology.
We hope that through this approach, users can personally feel the smooth experience and value transformation brought by ZK technology, understanding why this off-chain computation and on-chain verification model can make applications both secure and efficient. More importantly, we want to encourage more users to become promoters of the ecosystem: as more users understand and actively demand that the products they use adopt this mechanism, the entire industry's innovation momentum will also be activated in reverse.
Rhythm BlockBeats: Besides this event, what other plans does Brevis currently have? Can users participate in them?
Michael: We will continue to launch participatory ecosystem plans, including promoting the launch of a decentralized proof node network to form a more open proof and verification infrastructure. Relevant plans will open participation channels for the community and ecosystem partners, with specific participation methods and timelines to be announced later.
Rhythm BlockBeats: From a longer-term perspective, what kind of infrastructure do you hope Brevis will become? What role does it ultimately want to play in the Web3 world?
Michael: Our vision is to make Brevis the infinite computing layer of Web3, becoming the source of globally trustworthy computing. Similar to how Ethereum reconstructs the trust logic of assets and finance through smart contracts, Brevis aims to reconstruct the trust logic of computation: no longer choosing between self-computation and trusting others' computation results, but ensuring the safety and correctness of results through verifiable computation without repeated execution.
Looking to the future, on-chain applications will evolve from simple transaction execution and asset transfer to smarter, more complex systems, encompassing advanced algorithms, data interaction, personalized experiences, and proof of social behavior. To achieve large-scale adoption, it is essential to have both powerful computing capabilities and comprehensive security. We want to make this proof-based computation universal, user-friendly, and low-cost—allowing every Web3 application to naturally embed this capability without reinventing the wheel.
From an architectural perspective, we hope to add a new layer of trust—an infinite computing layer—on top of the consensus layer, data layer, and execution layer. It will become the fourth layer of infrastructure for blockchain, providing unified computational trustworthiness and verification capabilities for the entire decentralized world. This will allow all intelligent, complex, and even cross-chain computations to be verified and used under a zero-trust premise. This is our understanding of infinite computing and Brevis's long-term mission.
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。
