zkSync VS Solana 单链与多链生态的未来之争
作者：Haotian，独立研究员 来源：X，@tmel0211
有朋友问我，如何看 @zksync 和 @solana 两位创始人关于“单/多 链生态”未来的争论？先说结论：嗯，饼很香，你们都说的对！
1）zkSync的未来Hyperchain多链时代是解耦以太坊单Monolithic链的一种堆栈化扩容解决方案。它把以太坊视为结算层，以ZK技术为通信验证为核心媒介，再以layer3为应用流量入口层，最终形成一个layer1结算+layer2互通操作性+layer3应用多模块Stack战略。
总的来说，Hyperchain的未来确实值得期待，既避开了layer1主链的性能缺陷，也抓住了layer2的可交互操作的流动性聚合通信网络核心，layer3应用链想象空间也无限大。
不过，目前layer2处于构建基础Sequencer+Prover系统的infra阶段，layer3还处于萌芽时刻。未来虽然很美好，但距离还很遥远。
2）Solana驱动的超级应用时代则是基于纯单体链的一种落地方案。在这个语境下，Solana的高并发性能已经可以支撑多应用同时运转，无论通信、可交互操作性、Gas消耗等都比肩甚至超越以太系的Stack堆栈策略。
这也没错，Solana本就是为未来超级区块链设计的单体链，其接近“光速”的同步状态，以及单一原子状态机中处理多并发交易等链原生性能设计远超以太坊，Solana最初目标就是一种通向区块链应用未来的架构。
只不过现在整体行业还停留在基础infra的构建阶段，超级应用的需求并未大规模落地，Solana的高TPS、高并发等性能也尚未经受真正的考验。
3）应该说zkSync和Solana两位老板的论战并不在同一个纬度，二者都规划了一个应用驱动的超级链未来，无非是实现的路径不同。zkSync是以太坊的模块化堆栈链，利用零知识证明技术，来提高交易速度和降低成本；而Solana则是一个高性能的单体区块链，通过其独特的共识机制和系统架构来实现高吞吐量和低延迟，旨在成为支持大规模应用的基础设施。
无论是Stack多模块组合方式还是单体闭环生态实现方式都有很大机会。在愿景都尚未实现前，谁也没资格Diss谁。
4）至于toly所说的 ZKP存在的数据库热点问题确实存在，让智能合约短时间接受大量请求难免会出现拥堵或处理延迟，这是Stack多模块组合必然会面临的问题，本质上扩展链只是分离了共识层、执行层、应用层等堆栈，基于可交互操作性来扩容交易，仍然存在数据库资源优化和分配问题。
这一点似乎Solana原生链更有优势一些，甚至toly会骄傲的认为TPS是Solana数据优化和并发处理之后的必然结果，都不需要刻意提。然而，zkSync却认为使用ZKP可以通过异步交易处理来提升处理能力，ZKP技术本身就为解决原生链无法并发交易的问题而设计。
5）最后toly暗指ZKP的异步处理虽然会促成并行处理，但其可能会导致去中心化全节点参与的弱化，因为大量高效处理ZKP肯定集中化的节点效率更高。
不过，Alex并没有直接回答这个问题，说了句处理1M TPS的全节点可能需要一个计算机集群，但zkSync更想在手机上以轻量化的方式验证。表明zkSync追求的是一种即便在设备能力有限的情况下也能参与验证过程的去中心化模型，尽管这可能面临处理高吞吐量（如1M TPS）的挑战。俩人的争论和“过家家”一样，有点意思。
貌似，Alex经常和同行有这样的“双方似乎都有道理”的技术论战，或许是西方开发者社区的一种互怼营销策略吧，仅此而已。不然，吃瓜群众们还能怎么看呢？
The author is an independent researcher. A friend asked me what I think of the debate with the two founders about the ecological future of single-chain and multi-chain. Let's start with the conclusion. Well, the cake is delicious. You are all right. The future multi-chain era is a stacked expansion solution to decouple the single chain of Ethereum. It regards Ethereum as the settlement layer, takes technology as the communication verification as the core medium, and then thinks that the application traffic portal layer will eventually form a settlement and interoperability multi-module strategy. Generally speaking, the future is really worth looking forward to, which avoids the performance of the main chain. Defects have also been caught, and the imagination of the interoperable mobile communication network core application chain is infinite, but it is still in its infancy at present, and the super application era driven by pure monomer chain is a landing scheme. In this context, the high concurrency performance can support multiple applications to run at the same time, which is comparable to or even beyond the stack strategy of Ethernet, regardless of the interoperability consumption of communication. The wrong book is the single chain designed for the future super blockchain. Its synchronous state close to the speed of light and the design of the chain's native performance of handling multiple concurrent transactions in a single atomic state machine far exceed that of Ethereum. The original goal is a future architecture leading to blockchain applications, but now the whole industry is still in the basic construction stage, and the demand for super applications has not been put to the real test. It should be said that the debate between the two bosses is not in the same latitude. Everyone has planned an application-driven super-chain. The future is nothing more than the realization path. The difference is that the modular stack chain of Ethereum uses zero-knowledge proof technology to improve the transaction speed and reduce the cost, while it is a high-performance single blockchain. Through its unique consensus mechanism and system architecture, it aims to become an infrastructure supporting large-scale applications. Whether it is a multi-module combination mode or a single closed-loop ecological realization mode, there is a great chance before the vision is realized. No one is qualified. As for the hot database problems mentioned, it does exist. It is inevitable that there will be congestion or processing delay when smart contracts accept a large number of requests in a short time. This is an inevitable problem faced by multi-module combination. In essence, the extended chain only separates the consensus layer, the executive layer, the application layer and other stacks to expand the transaction based on interoperability. It seems that the original chain has more advantages, and some even proudly think that it is data optimization and concurrent processing. However, it is believed that the technology itself is designed to solve the problem that the native chain can not trade concurrently. Although the asynchronous processing implied in the end will promote parallel processing, it may lead to the weakening of decentralized participation of all nodes, because a large number of efficient processing will definitely make centralized nodes more efficient, but it does not directly answer this question. I want to verify on the mobile phone in a lightweight way, which shows that we are pursuing a decentralized model that can participate in the verification process even if the equipment capacity is limited, although it may face the challenge of dealing with high throughput. The debate between the two people is as interesting as playing house, and it seems that there is such a technical debate between the two sides that seems to be reasonable. Perhaps it is a mutual marketing strategy of the western developer community, that's all, otherwise what can the people who eat melons think? 比特币今日价格行情网_okx交易所app_永续合约_比特币怎么买卖交易_虚拟币交易所平台
注册有任何问题请添加 微信：MVIP619 拉你进入群
打开微信扫一扫
添加客服
进入交流群
1.本站遵循行业规范，任何转载的稿件都会明确标注作者和来源；2.本站的原创文章，请转载时务必注明文章作者和来源，不尊重原创的行为我们将追究责任；3.作者投稿可能会经我们编辑修改或补充。
