Today, I would like to talk with you about a project on the public chain track – sui.
The public chain track can be described as the most basic and bottom-level cornerstone of the web3 industry. It is no exaggeration to say that the public chain is the creator of the entire industry, and the emergence of Bitcoin is the genesis chapter of the opened industry. The development of the public chain is extremely difficult. The development time is often one or two years, and the long time is two or three years. Although there are no other sectors crowded, the competition is equally fierce.
In the 20-year bull market, we have seen many public chain projects springing up like bamboo shoots, solana, polygon, Avex, fantom, BSC, Polkadot …
When any public chain appears in the public eye, it declares itself to be an “Ethereum killer”. (But Ethereum is getting stronger and stronger, and the ecology is getting more and more prosperous. In August, the merger of beaconchain and layer1 will be carried out, and it will be officially transferred to the POS algorithm. Could it be that everything that does not kill Ethereum will make Ethereum more and more getting stronger?)
Returning to the topic, although each “Ethereum killer” has various technical solutions, consensus algorithms, economic models, and governance, in fact, all problems point to one core – the “impossible triangle” problem (throughput, security, decentralization) .
We would intuitively think that what limits the development of the integrated industry is the throughput problem (also known as the performance problem) and ignores the other two “corners”. Last year’s star project solana had extremely high performance, but it was actually based on the premise of sacrificing the foundation of decentralization and security to a large extent. During the development period, there were often downtime incidents and were criticized by the industry.
Therefore, in my opinion, the best public chain solution is actually to greatly improve performance without sacrificing decentralization and security as much as possible. This is the big logic of my view of public chain projects.
With this big logic in mind, let’s take a look at the star team project – sui, which is about to be launched on the testnet.
Let’s briefly introduce the background of the sui project.
I don’t know if you still remember, the Meta (former Facebook) company, which is now all in the Metaverse, launched a project called Libra (later renamed Diem ) in 2019. Due to supervision and other reasons, the project was stagnant, and the cost was spent before and after. 20 million, because it has not been put into practical applications, most of the money is invested in basic research and development, so it has accumulated a large amount of technical foundation. After the project was stopped, the core developers of the project forked out two entrepreneurial teams, Aptos and sui. Because Web3 has no shortage of hot money, Aptos quickly got an investment of 200 million US dollars (valued at 2 billion). The sui team is also actively raising funds and will launch a test chain soon.
Top-notch project development teams are different from the rest of the world by sticking to strong technology rather than tinkering with existing solutions.
performance! performance! performance!
Let’s take a look at sui’s technical solutions in detail. Let’s talk about how sui solves the performance problem of one of the cores of the public chain mentioned above.
Before talking about this issue, let’s look at another thing:
When the blockchain was originally designed, the data structure was defined as a linear chain structure. In order to obtain rewards, miners competed for the right to produce blocks, spliced the latest blocks to the latest chain, and changed the entire system database. state. Therefore, the working mode of the whole system is single-threaded.
Anyone in the computer industry knows the limitations of single threading. Compared with “multi-threading” (Xiaobai can also be simply understood as parallel), the speed is slow and the resource utilization rate is low. Different from other public chains, the sui team designed their public chain to be a multi-threaded block generation method.
Intuitively, this method is simple, clear, efficient, and mature in traditional computer technology. It is impossible for other public chain designers not to think that the reason why it has not been implemented is because there will be many derived problems in this method. It affects the ecological security of the entire public chain.
The most direct example is the “double spend” problem. Suppose a public chain adopts a multi-threaded block generation method. When a user initiates two transactions at the same time, they are accepted in both threads at the same time. If both threads are packaged by miners and confirmed by block generation, the money will be Spend twice. If the transaction information of the two threads is compared in order to solve the double-spending, it will cause the comparison cost or the risk of rollback, but increase the work cost of the miners.
And how does the sui chain realize multi-threaded work on the premise of avoiding security problems?
First, he divided transactions into two types:
1. Independent transactions
2. Dependent transactions
For example, suppose we have USDT in our wallet and some ETH in the compound . I want to do the following:
1. Change the U of the wallet to BTC in Uniswap
2. Take out the ETH in the compound. We will find that operation 1 and operation 2 here are independent of each other. This is the “independent transaction” in the sui chain.
If I want to withdraw the ETH in the compound, the withdrawn ETH will be exchanged for USDT in uniswap.Here we will find that these two transactions are related and have dependencies, and this is the “dependent transaction” in the sui chain.
After understanding these two concepts, you can also have a good understanding of the multi-threaded operation mode of sui. First of all, they will execute independent transactions in different threads, and execute dependent transactions in the same thread, which is a bit like “Transaction” operations in mysql.
Moreover, the consensus algorithms adopted by the two transaction types are also different to ensure the security of the two types of transactions.
If multi-threaded work can be achieved, the performance of the public chain can be greatly improved (according to team members, sui runs on a Mac notebook and handles simple transactions, and the TPS can reach 120,000 per second). As for whether it will bring other problems, I have a question here, it will take time to test, after all, sui is the pioneer of this solution.
Move contract language
In addition to multithreading, sui has also been upgraded at the language level.
They invented a new contract language – move, which is a new language paradigm based on rust and improved. Here is a background for those who do not understand technology. Rust is a language that can run at a speed comparable to C, and can perform full-stack development. It is a language recognized by the industry with great potential. Also, unlike Solidity’s account-centric storage model, move’s storage is designed around objects. There is a mapping relationship between objects and addresses, which will not change under normal circumstances. Under special requirements, an object can also be shared by multiple addresses. Functions in the Move contract can also receive objects as input and output.
Move language introduction link: https://medium.com/mysten-labs/recap-sui-ama-move-programming-language-with-todd-nowacki-damir-shamanaev-bfc24cc7f0f
Due to the dynamic changes in the demand of the entire blockchain network, the gas fee is often in an unstable state, and even in Ethereum we often see the phenomenon of malicious competition raising prices.
In response to this problem, sui has a new design.
First, sui defines an epoch, each epoch is 24 hours long. At the beginning of an epoch, all miners will vote on the gas fee for this epoch. After the vote, the entire epoch will be executed according to the originally voted gas until the next an epoch. cycle back and forth.
You might ask, what if network demand skyrockets during this period?
At this time, sui activates the inflation system to issue more sui as rewards to motivate more miners to do computing work. Because it is a multi-threaded work, the new miners will not affect the original miners. Note that the rewards obtained by miners here do not come from users, but from the issuance of the chain itself.
Finally, let’s talk about the economic model of sui. The total supply of SUI tokens is 10 billion. It will be distributed among the founding team, investors, public sale, sui foundation and miners. Here’s a point of view. The success of the sui chain is another matter. Judging from the number of tokens he has issued, he has a relatively grand plan and ambition.
Sui’s token has 4 functions:
- Pledge/Security (it can be seen that sui will adopt the consensus algorithm of pos to produce blocks)
- Transaction fee (gas fee)
- Unit of account/medium of exchange
And sui will try to use the economic model to solve the problem of “state explosion”. What is state explosion?We all know that the data of the blockchain is a chain structure that cannot be tampered with, so the amount of data will increase over time, and the cost for miners to access nodes will also increase. For example, the current Ethereum The amount of data is 810G. There will actually be a lot of garbage data here, such as abandoned smart contracts or data. In order to solve this problem, sui adopts the method of mortgage storage. In addition to the necessary data, users need to use sui for mortgage in order to store any data on the chain. When the data is not needed, the token will be taken In return, the sui chain will release the data, so that all necessary data is left on the chain.
1. The sui public chain team comes from the former libar team, and the lineup is luxurious
. 2. By defining the transaction type, multi-threaded block production is realized.
3. Try to solve the problem of unreasonable gas and “state explosion”.
4. The total issuance of tokens is 10 billion, and the specific allocation Details
5. Use the new contract language Move
6. Will be on the test chain soon, participate in the node test, you can get early bird rewards
Finally, let me talk about my overall opinion on this project. Due to the particularity of public chain projects, the development difficulty is extremely high, the engineering difficulty is extremely high, coupled with security issues and other factors, it is difficult to put it into production in the short term. The sui team has made a lot of improvements and innovations at the bottom of the public chain. The theory and ideas are really refreshing, but the effect after the actual investment is not only determined by one factor. For example, the multi-threaded block generation of sui has security or other factors. Derivative issues, such as whether the new move contract language adopted by sui can be recognized by the market, and whether developers are willing to migrate on the original basis, are unknown. I will continue to pay attention to how sui will develop in the future, let us wait and see.
Posted by:CoinYuppie，Reprinted with attribution to:https://coinyuppie.com/this-ethereum-killer-is-not-too-sui/
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.