区块链项目：ZILLIQA与以太坊的分片那点事

当今传统区块链面临的核心问题：扩展性问题

所有区块链交易所面临的最基本的需求就是交易处理能力：传统区块链处理每秒7个，visa是8000个，自从分片概念提出以后，传统区块链以太坊和后起之秀小鲜肉BILLIQA便同时开始在分片技术的实现上走上一条不归路。

目前公有链主要面临两大扩展性问题：

1）吞吐量低:比特币和以太坊每秒处理小于10比交易，不能满足目前的需求

2）存储空间占用： 记账节点需要保存所有的历史交易记录，比特币已经从09年到现在已经用150G的

分片技术主要用于解决以上两个问题：

1）交易分片：各个分片分别处理部分交易，提高交易效率

2）存储分片：各个分片分别存储部分交易记录，减低单节点的存储压力

功能上的区别：

1）以太坊实现2种分片

2）Zilliqa只能实现交易分片

以太坊之分片

以太坊V神路演之后提出你可以把分片看成是一个区块链，上面有上百个不同的宇宙，每个宇宙都是独立的账户空间一个账户，一个合约的生成和发起都可以在不同的宇宙上，也可以在多个宇宙间发起一笔交易，而这个交易仅仅会对涉及到的宇宙产生影响。

            分片机制：

      主链--分层--分片

各个分片是独立账户，生成区块的头部，主链收集各个区块的头部，每个分片把区块头部发送到主链，然后生成主链区块。

这里的意思是主链只保存各个分片的区块头，并不保存具体交易。多个分片可以并行处理交易，分散主链压力，提升整个系统的吞吐量。分片节点保存与交易相关的历史记录，主链节点也不用包含分片的具体交易信息，从而大大降低节点的存储压力。

分片共识：pos

进度：3-5年

吞吐量：整体提升，主链没有

各个分片是独立账户，生成区块的头部，主链收集各个区块的头部，每个分片把区块头部发送到主链，然后生成主链区块。

ZILLIQA 分片

ZILLIQA的核心优势：

对于传统区块链莱说，所有挖掘者都争先恐后的解决一个数学问题，获胜者才能得到奖金，这是一个赢家通吃的游戏，这种竞争结构使得剩下的99%的失败者浪费了计算机算力和消消耗大量能源。

例如传统区块链有8000个节点，他们同时解决一个问题，而分片是分出10个分片，每个分片是800个节点，分别解决不同问题，竞争变成合作，计算机的耗能更加有效。随着挖掘者的不断增多，可以处理更多的问题，区块链可以实现更多的交易，随着运行效率的增加，整个网络的交易成本也会大大降低，挖掘者获利不变。

分片的机制：运行在同一个空间

全部矿工随机分配 筛选出一个boss分片+n个工作分片，每个工作分片分出y个矿工，全网节点首先进行一次数学竞赛，最先胜出的m个节点组成boss分片。

剩余的节点进行第二次竞赛，最先胜出的n*m 节点分配到n个工作片中，每个分片都并行处理部分交易，形成子区块，发送到boss分片，Boss验证，打包，生成最终区块。无论是boss分片还是工作分片都工作在主链上，没有独立空间，每个节点都保存的所有的历史记录。

分片共识：PBFT

进度：测试网络已经发布

吞吐量：主链提升

优点：

节点之间是合作关系，而不像传统区块链是竞争关系。

ZILLIQA 代币（ZIL) 登陆 交易平台：

火币pro：huobi.pro

cate.io

idex.market

lola

区块链研习社精英群成员

2018.1.28

心痛小姐姐的辛苦汇总的币豪们，小姐姐接受eos,eth,btc以及各种币币打赏

谢谢比豪和未来的币豪们：

币币打赏地址：

ZILLIQA地址：

0x79d1280bc971dff6327d0a00360f23c507c25a33

BTC地址:

1PibSCEynK44Q9ef5b5WYurEb5y9dCd79y

ETH地址：

0xe9aab08d1e11b49648c768e1351879e53270c5f2

EOS地址：

0xe9aab08d1e11b49648c768e1351879e53270c5f2

QYB(区块链研习社代币）：

地址: QUmjuR8rzmbjCqunZTdjRfqFq1if873YKW

区块链研习社LOGO

加入研习社扫二维码：

最新消息：

https://www.techinasia.com/talk/singapore-based-blockchain-compete-visa

As crypto hasseen massive growth in 2017, the biggest platforms are running into fundamentalproblems due to scale (neither bitcoin nor Ethereum can support more than threeto seven transactions per second). It doesn’t matter what you do to the blocksize or what other fixes you use, blockchain is no competition againstsomething like Visa that processes 8,000 transactions per second.

To make mattersworse, “traditional” blockchains (bitcoin or Ethereum) don’t scale as theygrow. As more and more people enter the world of crypto, they will have to bidagainst each other to get their transactions in a block, while more miningpower will compete for the consequently bigger block rewards.

But wouldn’t itbe better if blockchains could scale their transactions as the ecosystemscales? Of course it would. But due to the fundamental setup of theseblockchains, there’s no real solution. The only option is to move most of thetransactions off-chain or into side chains (think the Lightning Network, forinstance). But isn’t the whole idea of a blockchain to have alltransactionson the blockchain?

Looking for change

A team from theNational University of Singapore (NUS) is challenging these assumptionswithZilliqa. Led by its CEO, Dong Xinshu, theZilliqa proposal has been on the table since 2015 but is only now ready for themarket. The team is mostly Singapore-based and draws a lot of raw talent fromNUS.

The startup is ablockchain infrastructure quite like Ethereum but structured in a way wherehigh levels of scale and speed are possible without the excessivecosts. The fundamental premise is that transactions scale while thenetworks scale—more miners equals more transaction throughput, making a moreuseful blockchain.

All this isachieved through a concept called sharding. In a traditional blockchain, allminers work on a very hard mathematical problem. They each compete to solve itfirst, and once that’s done, the winner gets a reward. It’s a winner-takes-allgame. This system works well but has a big downside: everyone who’snotthewinner wastes his time (or computer power, in this case). Due to the way thecompetition is structured, essentially 99 percent of the computer power usedgoes to waste.

Sharding breaksthe winner-takes-all dynamic and the cryptographic task (validating thetransactions) down into small segments or “shards.” That means that instead of8,000 miners competing to solve a problem, for example, these miners are automaticallysplit into 10 shards of 800 each, working on separate problems.

This way, youget computational cooperation rather than competition. More computer power isused fruitfully.

This setup leadsto a couple of interesting consequences:

As the network grows, the cryptographic    difficulty doesn’t scale. Rather, more shards get created, where more    miners can work on more problems. As an effect, more transactions come    through the blockchain.

Because of this, GPUs can do the mining rather    than specialized mining hardware.

We get way more cryptographic work done for the    same electricity output. It’s much greener.

Because everything runs much more efficiently,    the transaction costs throughout the network can be much lower while still    being equally profitable for miners.

Future andchallenges

Zilliqa has beenin development since 2015. In recent testing on an internal testnet of AmazonWeb Services (AWS) instances, the network breached 2,488 transactions persecond. That’s more than 100x more than bitcoin and Ethereum combined.

But that waspre-scale. If all Ethereum miners (~22,000) work on Zilliqa, that would allowfor about 15,000 transactions per second. That’s more than twice as many asVisa (the world’s largest payment processor).

While Zilliqawould be the first project bringing sharding technology to a big publicblockchain, it must be noted that sharding has some challenges too. Thetechnology still needs to be tested and vetted at scale. Sharding could alsopotentially introduce security challenges, as attackers could try to hone in onone shard rather than the entire network.

But for now, thepromise of blockchain linear scalability seems to be driving a lot ofexcitement.

On January 3,2018, the team released the much-anticipated “Durian” alpha sourcecode. Now, everyone can inspect the underlying technology, and developers cantest and improve it. With a stable proof-of-concept already down, the companyis announcing an ICO this month. The team announced they were going tolower their hard capfromUS$30 million down to US$20 million, confident that their roadmap could beexecuted with these resources.

The premise issimple: Whoever can build a stable, reliable blockchain infrastructure that cancompete at the levels of Visa and MasterCard can be a very substantial playerin the global economy.