On eth_deofs Day, @ drakefjustin verified the proof of @ brevis_zk Pico ZKVM on-site using the zklighthouse client - finalizing blocks without the need to re execute transactions, and shouted that the era of Gigagas ETH is coming. Undoubtedly, Brevis has once again been appointed by the Ethereum Foundation. What game is Brevis playing? After reviewing their latest ProverNet white paper and adding a demo demonstration, it was found that it is indeed different from other zk solutions: 1) In general, ZK proves that the market follows a single supplier model, but Brevis found an awkward problem in the production environment: for example, the off chain rebate process of @ Uniswap requires ZK to provide a transparent resolution method, the VIP rate of @ PancakeSwap needs to implement sub second level proof checks before each transaction, and the incentive allocation of @ eulerfinance, which requires a certain amount of time to batch process over 100000 addresses, has throughput requirements. According to normal logic, several different demand scenarios require completely different hardware configurations, proof system specifications, and optimization directions. If it is a single supplier, either they can only perform in-depth scenario optimization for temporary needs, or they can only provide a comprehensive solution for generalization. In this way, it is obviously difficult to meet the tricky differentiation needs in vertical scenarios; 2) The ProverNet solution that Brevis wants to do is very simple. It is to build a two-sided market, where the application side proposes requirements and professional Provers bid to accept orders, that is, TODA (True Reproduction Double Auction) mechanism. This "market-oriented" solution can naturally handle multiple heterogeneous proof types simultaneously, allowing applications with segmented scenario requirements to receive the most professional and proficient Prover for specialized matching services. In fact, it can be further improved. If encountering more complex tasks, different Provers can also break down the tasks for collaboration. For example, a zkVM proof may be divided into blocks by A, compressed by B, aggregated by C, and finally encapsulated by D. In short, professional division of labor and separation are necessary to meet more diverse market demands. The above. The logic is actually very simple, it is to shift from the mindset of "selling shovels" in the past to building a "market", from a purely centralized supplier service system in the past to a decentralized Marketplace. In the so-called Gigagas ETH era, what is needed is not a single tool service, but an infrastructure that can schedule Provers across the entire network.