7月11日晚,Ultrain超脑链受邀请在币安DEX Telegram 社群举办答疑大会(Ask me anything)。Ultrain 联合创始人Emma廖志宇、首席架构师沈宇峰在线和币安社区互动,八千多人参与活动,提交了数百条问题。本次答疑被同步直播至超脑、币安中英文诸多社区,覆盖10万人以上。




Darc: Good evening/morning everyone, we are very happy to have Emma Liao, Cofounder of Ultrain and Yufeng Shen, Chief Architect of Ultrain to be with us now and to answer any questions about Ultrain! first please let Emma and Yufeng give a quick intro of themselves and about the project.

主持人Darc: 大家好,我们很高兴能邀请到Ultrain联合创始人Emma廖志宇和首席架构师沈宇峰来此回答一系列关于Ultrain的问题。首先,有请Emma和宇峰给我们做一些自我介绍,并介绍一下Ultrain项目。

Emma Liao: Thank you Darc, hello everyone, i am very honored to be here and have the opportunity to discuss Ultrain with Binance’s and our community members. My name is Emma Liao, born in Beijing, grew up in the US and had spent 4 years in France and Germany for work. I am passionate about art, fashion and technology innovations from at young age. I have studied science for the whole time and pursued ph.d in Gamification at Columbia University. I also had a MBA degree in technology management and Finance. Born in a highly reputable general family in China, I did have the advantage to be close to many successful leaders and entrepreneurs early on. Later I was an important bridge for the east and west to help large cross-border Mergers and Acquisitions as head of China for US and European firms. in 2015, I started to pursue my dream of being an entrepreneur. I served as head of IOT at QIHU 360 then Managing Director at a 1billion dollar TMT fund. at the end of 2017, I met my talented partners Ray and William from Alibaba being the tech lead in data security and blockchain unit there, and we cofounded Ultrain together. Also I am an animal lover and very passionate about nature and animal conservation. 🙂

Emma: 谢谢Darc,大家好,很荣幸能有机会在这里与Binance和我们的社区成员讨论Ultrain。我叫Emma Liao,出生于北京,在美国长大,并在法国和德国工作了4年。我从小就对艺术、时尚和技术创新充满热情,一直在研究科学,曾在哥伦比亚大学攻读游戏化(Gamification)博士学位,并拥有技术管理和金融MBA学位。我出生在中国一个具有名望的大家庭,这对于我早期接触许多成功的领导者和企业家非常有帮助。


Ultrain is a world-leading public blockchain with high throughput, fast confirmation timewhile still being truly decentralized. Ultrain aims to build a general purposeblockchain infrastructure optimized to be the key support for future businessesand human interactions. Ultrain foundingteam consists of top cybersecurity and blockchain experts from Alibaba, Google,IBM, and an globally well known senior banker/investor with top business/politicalresources. Ultrain has already launched its main-net with brand-new smartcontract layer on Apirl, 15th 2019. It also has ~400 miners and 3 side chainsstably running. In a way we are the most decentralized and complete publicchain3.0 ecosystem globally. I will explain in more detail later.


Yufeng Shen: Hello everyone, my name is Yufeng Shen. I am theChief Architect of Ultrain. I graduated from Zhejiang University and Universityof Alberta. I am the former AliOS senior tech specialist at Alibaba CouldOSunit, was responsible for In-Vehicle Infotainment (IVI), and WebAppapplication. I have many experience in App ecosystem and specialize inimproving performance of both software and hardware. Before that, I was theformer senior software engineer at Google, responsible for the Linux kenellayer structure and Touch subsystem of OS Framework layer, entire Touchsubsystem of Google’s first high performance laptop ChromeOS.

沈宇峰:大家好,我叫沈宇峰,是Ultrain的首席架构师。我毕业于浙江大学和加拿大阿尔伯塔大学。我是前阿里巴巴CouldOS部门的AliOS高级技术专家,负责车载操作系统和WebApp应用。我在App生态系统方面有很多经验,专注于提高软件和硬件的性能。在此之前,我曾担任Google的高级软件工程师,负责ChromeOS上 linux 内核层驱动和 OS Framework 层的 Touch 子系统以及Google首款高性能触屏笔记本 ChromeOS Pixel 的 Touch 全栈模块。

Darc: Who do you view as your peers in the field?


Emma: Worldwide there are only two similar projects to us: DFINITY and Algorand. We have had a faster realization and we have done more than Algorand in terms of realizing the consensus layer, the smart contract layer and privacy layer via our own customizable zero knowledge proof solutions.

We have landed many business customers using our chain, and have had 400 fair and equal supplier (miners) to construct an open, decentralized and randomly selected mining network. Our intention is to have 1000 decentralized nodes by the end of 2019 and will be open sources and support PC mining by the end of 2019, which makes us a truly decentralized public chain network from the supply side.

People may feel that some private chain and constellation chain solutions are our competitor, such as hyper ledger from IBM and other BAAS providers from internet giants. However we believe in the world of real blockchain as infrastructure for new applications, public chain is still a key to really lower the boundary and cost of trust. Chains with DPOS mechanism has fairness and security issues and people could argue the nodes is in few people's hands or in other world they are not decentralized. Ultrain provides a truly decentralized infrastructure with 3rd party nodes, Ultrain provide much more performant (faster, quicker confirmation time) to cater broader business demands than traditional public chain such as Ethereum, and Ultrain offers a most economical pricing point with a super user friendly product.



人们可能会觉得某些私有链等是我们的竞争对手,例如来自IBM的hyper ledger和来自互联网巨头的其它BAAS提供商。然而,我们相信,区块链作为新应用的基础设施,公链仍然是能真正拓宽信任边界、降低成本的关键。具有DPoS机制的链存在公平性和安全性问题,人们可能认为节点掌握在少数人的手中,并不够去中心化。而Ultrain用第三方节点提供真正去中心化的基础设施点,具备更高性能(更快的确认时间)以满足比传统公链(如以太坊)更广泛的业务需求,并且还会提供最经济划算的定价和超级用户友好的产品。

Darc: Can you specify a more detailed comparison between Algorand and ULTRAIN please?


Yufeng Shen: We would like to elaborate mainly in terms of the similarity and differentiation from our consensus mechanism.

I.Algorand adopts the method of private election, indicating that each node can decide whether it will be elected to propose. Such an approach provides better privacy and overall security, since the nodes participating in consensus only reveal their identity at the broadcasting stage, thus eliminates the possibility of attacks. The disadvantage, however, is that the number of nodes selected during the first-round is execution through probability. If the parameter setting is too small, it leads to a higher chance of generating empty blocks. While, setting the parameter to large will leads to higher bandwidth consumption (because larger parameter results in too many proposer nodes, producing a large flow of messages within the network that burdens the system. The TPS performance is directly proportional to the bandwidth executed during the first round of consensus, the system’s TPS should be 1/n while the N representing the number of nodes.


I. Algorand采用私人选举的方式,每个节点可以决定是否选举提议。这种方法提供了更好的隐私和整体安全性,因为参与共识的节点仅在广播阶段显示其身份,因此消除了攻击的可能性。然而,缺点是在第一轮期间选择的节点数是通过概率执行的。如果参数设置太小,则会导致生成空块的可能性增加。然而,将参数设置为较大将导致更高的带宽消耗(因为较大的参数会导致提议节点过多,从而在网络中产生大量消息,给系统带来负担,TPS性能与在此期间执行的带宽成正比。第一轮共识,系统的TPS应为1 / n,而N代表节点数)。

Ultrain uses a public random number approach, which is also used in Cardano, Dfinity, and Ethereum’s next-generation networks. The difference is that Ultrain uses random number voting based on VRF (Verifiable Random Function), and the voting members are selected from the candidates participating in the consensus committee. The random number submitted by all members of the voting process is constrained by the VRF function and the blocks generated in the previous round, so it is impossible to manipulate or control the random number in the end. In the meantime, the system rewards the voters and punish the non-voters to ensure the security and fairness of the random number mechanism. By adopting an encrypted decentralized algorithm in the second layer (off-chain), each node can join as a group to prevent malicious behavior and further ensure overall network security. Moreover, each round of consensus will limit the number of proposing nodes and voting nodes, ensuring that network bandwidth consumption can be predicted, and the overall performance of the blockchain is stable.

The existing Randao approach requires two rounds of consensus, submission first and reveals algorithm second. There is a possibility that the user can submit any random number and manipulate the final result. Dfinity or Cardano, which uses the encrypted decentralized sharing algorithm, their network performance depends on the underlying group size. If the group size is too small, the user will collude/manipulate the random number or even conduct malicious attach. If the group is too large, the internal communication consumption of the group is O(n^2), and the calculation amount would be O(n), making an enormous consumption on network bandwidth.


现有的Randao方法需要两轮共识,先提交再显示算法。用户可能会提交任意随机数并操纵最终结果。 Dfinity或Cardano使用加密的去中心化共享算法,其网络性能取决于底层的组大小。如果组太小,用户将串通/操纵随机数甚至进行恶意附着;如果组太大,则组的内部通信消耗为O(n ^ 2),并且计算量将为O(n),从而导致网络带宽的巨大消耗。

II. Secondly, Algorand needs a large number of nodes to hold tokens online in real time, users who are not online will affect the overall security of consensus, and even preventing the network from reaching a consensus. This is a natural problem brought about by the combination of decentralized PoS and BFT, because BFT requires that nodes to hold a stake in PoS in order to participate in the consensus, and any nodes that are not online is considered as“malicious node”. On the other hand, EOS relies on a small number of nodes that hold a large number of tokens to execute its consensus mechanism, but raise a new issue of being semi-centralized. The network is actually manipulated in the hands of several privileged users or limited nodes permitted to participate in the consensus.

Ultrain requires uses to collateral their token to participate in consensus, and each token holder can participate in the consensus by collateralizing a fixed number of tokens (stake), thus enabling everyone to participate in a fair network. And the collaborated token holder is incentivized for their online participants to keep the network running smooth and well maintained, thus ensuring the final consensus of a network to be secure and effective.



III. Thirdly, Algorand is unable to motivate voters and cannotguarantee consensus participation.The voters are importantmembers of the BFT consensus, as they guaranteed the consensus consistency ofthe entire network. However, since they are not qualified as a node, thoseparticipants cannot receive appropriate incentives. Due to a large number ofvoters, misappropriation of incentives will seriously affect the performance ofthe system. Each of the block nodes will face different sets of voters, makingit impossible to incentive through the traditional approach.

Ultrain uses delayedrewards approach for voters and lottery rewards as incentive plans so thatvoting nodes have sufficient incentives to participate in consensus, thusensuring the security of the consensus mechanism. At the same time, Ultrainuses BLS for aggregate signatures, which greatly improves the verificationefficiency of voters and conduct consensus at the lowest cost.



IV: Last but not least, security concerns. Ultrain also solved the issue of Long Range Attack (The attacker creates a fork from an earlier block, causing the existing node to hardly distinguish between the forks.) for the PoS network by introducing VDF (Verifiable Delay Function) and rejecting the “old blocks”.

On top of that, Ultrain has far more advanced smart contract layer and business friendly solutions that enables enterprises and individual developers for easy plug and play adoption. Algorand so far has limited smart contract capability and applications. We believe the two projects are at very different goals therefore we may grow a different kind of ecosystem on top of our chains.



Darc: what is the current status of Ultrain tech/mainnet ?


Yufeng Shen: We have launched our mainnet back on April 15th, 2019; It was world’s first permissionless public blockchain network with fully sharding capability; it now has one main chain + 4 shard chains with total about 400 nodes, of which 90% are voluntarily joined, forming a truly decentralized network. Every one is welcome to stake and join our network freely. We expect by the end of 2019 the network will reach around 1000 nodes located worldwide. The network status can be checked out at https://explorer.ultrain.io/ultrainio.


In terms of realized technology features, we have finished R-POS consensus with about up to ~1000TPS and ~10s confirmation time in single chain; Ultrain dev suite (web/mobile wallet, JS dev tool, WASM VM runtime) are all ready for use; Sharding was done along all computation/storage/network dimensions making unlimited horizontal scalability possible and now the total network throughput can reach about ~4000TPS. In terms of security, we have done reshuffle of node-to-shard assignment, merkle proof and BLS aggregating signature based lightclient for main/shard chain communication; Also world-state snapshot based chain state sync makes it almost instant for node to migrate from shard-chain to shard-chain. More details about the technology can be found at https://ultrain.io/paper.

在已实现的技术特性方面,我们已经完成了R-PoS共识,单链约有1000TPS和10s确认时间。Ultrain开发套件(网页/移动钱包,JS开发工具,WASM VM运行时间)都已经可以使用。我们在所有计算/存储/网络维度上进行分片,从而实现无限的水平可扩展性,现在总网络吞吐量可达到约4000TPS。在安全性方面,我们已经完成了对主/分片链通信的节点到分片分配,merkle证明和基于BLS聚合签名的lightclient的重新洗牌。此外,基于世界状态快照的链状态同步,使得节点几乎可以立即从分片链迁移到另一个分片链。有关该技术的更多详细信息,请访问https://ultrain.io/paper。

Next the tech team will keep investigation on network security (e.g. VDF augmented consensus mechanism against long range attack; general purpose fraud proof for compromised shard-chain mitigation); data privacy (e.g. general purpose zero-knowledge proof); and embracing cross-chain ecology (e.g. participating in Binance Dex 2-way token bridge; Polkdot parachain bridge).

接下来,技术团队将继续对网络安全进行调查(例如,VDF增强了针对远程攻击的共识机制,得到了针对受损的碎片链的通用欺诈证据)、完善数据隐私(例如通用零知识证明)、并深入使用交叉链生态学(例如参与Binance Dex双向标记桥、Polkdot平行链桥)。

Darc: What is the business model of Ultrain?


Emma: Ultrain has a mature business model designed with sustainability in mind. We have computational power Suppliers (miners) and demand Customers (Dapps developer or Enterprise customer) in our ecosystem, our business model is: the Customers has to buy UGAS to build and use our public chain service, for every 100 dollar the customer’s purchasing on chain, 80 dollar will be distributed to our Suppliers, 15 dollar will benefit back to the Customer ecosystem and the community, and 5 dollar will be given to Ultrain tech team for tech support. You can imagine we are the decentralized android + amazon cloud service.

To use our much more friendly and enhanced operating system its free, and to run your business logic, aka applications or enterprise usage, we charge ugas for the computing fee. So the more customer we have, the bigger our ecosystem is, the higher conniption of UGAS will be, and higher demand we have for the Suppliers(miners) Our miner has to deposit 42000 ugas to participate the mining network, currently our ROI is extremely high. And since our product is Iive and ready for use, and we share a great customer pool and a promising pipeline, and more individual developers will join us for a better product and much cheaper gas fee (1/30 of what EOS and ethereum is charing) we have a very positive outlook of creating our sustainable business ecosystem.

Emma:Ultrain拥有成熟的商业模式,其设计充分考虑了可持续发展。我们的生态系统中有具备计算能力的供应商(节点)和需求客户(DApps开发商或企业客户),我们的商业模式是:客户必须购买我们的通证来建立和使用我们的公锁服务,客户在链上购买的每100美元,80美元将分配给我们的供应商,15美元将回馈给客户生态系统和社区,5美元给Ultrain技术团队以获得技术支持。你可以想象我们是去中心化的android +亚马逊云服务。

为了使用我们更加友好和先进的操作系统,并运行您的业务逻辑,我们收取我们的通证作为计算费用。因此,我们拥有的客户越多,生态系统越大,我们对供应商的需求也越来越高。由于我们的产品已经面世并且可以使用,我们共享一个巨大的客户群体,更多的个人开发商将加入我们,以获得更好的产品和更便宜的服务费用(EOS和以太坊的1/30) 。我们非常积极地在创建可持续的商业生态系统。

Q5: What is the potential user base and market opportunity?


Emma:Ultrain is a public blockchain3.0 project. Ultrain’s vision is in the future to combine blockchain, AI andIOT technologies to be the infrastructure of programmable business society foreveryone’s daily business use support. Ultrain wants to help apply blockchaininto traditional businesses and create values for their business scenarios,whatever is supply chain, finance, e-commerce or entertainment. We also believepublic blockchain can create new way of transactions and business scenarios, sowe can support and incubate many new applications in commerce and entertainmentand other fields.


Ultrain from a business modelperspective is B to B to C and B to D toC models at the same time. We have enterprise solution unit and developersdevelopment and service unit to cater the two different type of customers. Wealso have our own app - UltrianOne to directly engage our miners and communitymembers and more.

Via our strong businessecosystem pipeline, our incubated applications with top traffic internetcompanies, and growing developers community, we believe our adoption users basewill be a very large group of people. Our goal is to have 100k users of UGAS bythe end of 2019, and by 2020 we will reach to 1 million users, by 2021 we wishto reach 10 million users.

从商业模型的角度来看,Ultrain同时是B2B2C以及B2D2C模型。我们有企业级的解决方案和开发人员开发和服务单元,以满足两种不同类型的客户。我们还拥有自己的APP - UltrianOne,可以直接吸引我们的节点供应商和社区成员等。


After launching our mainnet onApril 15th 2019, there are already 5 businesses customers on our chain, includeYoho! (The Chinese version of StockX, the largest trend fashion e-commerce inChina), Unitopia (a leading video game company listed in China A-Share), GreatResources (a new energy company listed in A-Share), UC Express (A leadinglogistic company in China), and 8win.com (the leading live betting company inChina and UK). Meanwhile, we have a strong pipeline of leading enterprisescustomers with more than twenty companies from various industries, includinghealthcare, ad tech, retail, manufacturing, supply chain, energy,media&entertainment, and fin tech industry. The pipeline customers are allglobal fortune 500 companies, No1 or No2 players in their vertical, customerswho owns deep understanding and resources in their own industry.

Ultrain intends to leverage itsbusiness and tech resources to nurture landmark and lighthouse applications onpublic blockchain for good example and mass adoption to the blockchainindustry.