Exclusive report On May 28, the Blockchain Technology Development Forum (GBF) with the theme of "Technological Innovation and Integration" was held at the Guiyang International Ecological Conference Center.
In his keynote speech on "How to break through the key technical bottlenecks of the blockchain", Long Fan, an assistant professor at the University of Toronto and co-founder of Conflux, said: the biggest bottleneck restricting the development of blockchain technology is the consensus algorithm, but it is necessary to build a scalable Blockchain network, we need to solve all technical problems; there are two main bottlenecks for technological breakthroughs, namely "how to automatically write external data into the distributed blockchain system" and "how to ensure sufficient distribution without sacrificing security Improve the efficiency of the blockchain system under the premise of security".
The following is the transcript of the speech edited by Odaily:
Hello everyone. It's a pleasure to be here.
Tell me about my own journey. I was engaged in research work five years ago, and my research directions are distributed systems, system security, and programming languages. It is also because blockchain technology is related to many things I research, so I started to enter the related fields of blockchain.
I have a research group at the University of Toronto that specializes in blockchain-related research. I believe that blockchain is a revolutionary technology because I think the Internet solves data for direct information communication between people, and blockchain The chain solves higher-level needs, that is to say, in addition to transmitting information between people, many times everyone gathers together to complete a task, reach a consensus, and move forward for the same goal.
Reaching a consensus requires the use of blockchain technology. This is why I think blockchain is a revolutionary thing, and why I am willing to invest in it. In fact, there are still many controversies about blockchain technology. Not everyone holds the same opinion as me. My colleague from the University of Toronto does not like blockchain technology. He put forward a point of view: blockchain is Las Vegas. casino. I said why? He replied why not? Most of the users of most public chains in the blockchain exchange their earned Canadian dollars, renminbi, and U.S. dollars through a certain channel to find digital currency exchanges and exchange them for so-called chips. In fact, what they can do is very limited. Now all blockchains are played. From the very beginning, everyone has experienced the ICO boom. He said that you can see that many projects in the ICO boom turned out to be unreliable, so it is essentially a drumbeat to spread flowers. gambling game. But now many public chains, such as US, play gambling games.
When I was discussing this matter, I felt speechless. How could I refute him? I can't say that the things I do are very meaningless things, or things for gambling. I thought of this way of rebuttal. I said, you know that I am from Beijing. There is a very painful thing in Beijing. What is it? The air is not good, there is smog, especially not very good for the respiratory tract, every time there is smog, the throat feels very uncomfortable. We can use blockchain technology to do beneficial things, we can buy smog insurance, we use our favorite smart contract, all participants can contribute a little money to this smart contract, and then form a pool, when the smog When it happened, we bought people who had smog insurance for the day from this pool. Whether this is an innovation for the economy, of course, there are many details in it. Let's put aside these details first. When I was about to tell my colleagues this example, I suddenly realized a very important problem, that is, I can indeed solve such a smart contract in my favorite public chain. However, how to obtain the PM2.5 value in Beijing yesterday and the day before yesterday cannot be solved by any mathematical formula or data, because this is what really happens in the real world, and it cannot be obtained through calculation.
I will think about it carefully later, all blockchain-related and truly integrated with the real economy will encounter such problems, because if we want to use blockchain technology to empower the real economy, it will inevitably involve You need to know what happened in the real world, what happened in the real economy. Now the public chain is very similar to a closed Disneyland. The only entrance for everyone is through the digital currency exchange. After entering this park, because it is isolated from the real world, there is no other combination with the real economy except for playing in this park. way, way of empowerment. This is actually what I think is the biggest problem with public chains or blockchains now.
Many people are also proposing solutions, a popular one recently, there are many public chains saying that we use super nodes, I don’t know what kind of nodes are called super nodes, you have enough computers and enough channels It is called a super node, and any problem can be solved by the super node. As for whether it is more than 20 super nodes or more than 40 super nodes, it has become the solution of the public chain.
What I want to emphasize here is that the essence of the public chain or the public chain is actually to tie every participant, every individual, so-called linearly together, through a huge consensus agreement. If the data generated by the chain is verified by the big data people, then its data will not be tampered with, and this trust is guaranteed by the participation of the majority. We often hear that the blockchain can do this, that, and solve privacy. In fact, the blockchain only solves one problem in the end, which is the problem of trust. Because apart from these things themselves, anything can be done more efficiently.
Search engines like Google and Baidu may have millions of visits per second, while centralized servers like Alibaba and Taobao Double Eleven can process hundreds of thousands of transactions per second. Now, the fully decentralized blockchain technology can have dozens of orders of magnitude. What we want to use the public chain to solve is actually to tie all of them together, so that everyone can form a more powerful performance. That is to say, for the public chain, if we say that we are better decentralized, then the result of this public chain will be easier in the end, and in the end we will decentralize, we find 21 super nodes, we find For limited people, no matter whether this matter is through voting or what kind of method, the value of the data generated by this network and the value of trust will be discounted. In the end, you may think why I don’t just use a subject. It is very important to find a decentralized server instead of a trusted centralized server, which needs attention.
In my opinion, there are two main bottlenecks in the current technological breakthrough of the decentralized public chain. On the one hand, how to enable external data to be automatically and orderly written into the distributed blockchain system. Decentralized oracles, or we want to have highly reliable institutions to participate, and even the hot cross-chain is also part of it. In addition to the mechanism we need, the processing power at the bottom of the blockchain is also very important, because how much data do we actually have now? And if so much of the real world wants to be written on the blockchain, the current performance of the blockchain cannot be carried. You can say that it is highly scalable. We need to reduce the cost of writing to each blockchain. Only when we reduce the cost of the blockchain to a low enough level, will more applications be written on the blockchain.
To use an analogy, some people said before that we want to do taxi-hailing. Taxi-hailing is a good idea, because now we are decentralized, but what we consider is the technology we already have, the cost of each transaction we initiate, or It is said that the cost of the transaction may be a few yuan, four or five yuan, and I am now building a taxi-hailing platform on the blockchain. I called a car, sent my location, 20 yuan, and then the driver reported his location, 5 yuan, we successfully matched, 5 yuan, and the last order was 30 yuan. This is completely unrealistic, because your transaction cost is too high, but it does not mean that these are impossible. If we can reduce each value of the blockchain, this thing may become possible, if it is not five yuan Money, but five cents, this matter has a lot to do.
My colleagues and I are studying how to do this without sacrificing centralization and security. You may talk about various issues such as triangles. In fact, our view is that we are far from reaching a certain level. The limit that a single computer can handle, the performance of the current blockchain is actually limited by the consensus protocol. Of course, if you want to improve the overall performance, because of the technical bottleneck, the final technical processing capacity is determined by the shortest board, which must first process these transactions, involving network, storage, computing and so on.
If we go further, the P2P network layer, in fact, our current broadband can reach 2M, and the P2P network needs ten forwardings. We may have 3000-6000 transactions per second. We all use storage now. Every blockchain The read and write will be enlarged into dozens of SSD operations, so the final support is 2000-10000, depending on how you do it, the calculation can be 4000-8000, which means that our current hardware can fully support it, which is completely better than Decentralization can support much higher levels. This is what we can achieve without sacrificing security such as sharding or hyperlinks. We will eventually practice it and solve many technical problems.
What we have been doing is Conflux. We propose that the blockchain will eventually become an underlying infrastructure. It is a process in which valuable data is uploaded to this chain. We interact with each other, trust and share each other. Now the test has been completed. Yes, we hope to push forward our technology, decentralization and security, to reach thousands of transactions.
Our core team is led by Professor Yao Qizhi of Tsinghua University. There are distributed scientists, many of whom have overseas study backgrounds. In the end, our breakthrough consensus algorithm means that we have solved the high-speed block, so that the network under the high-speed block will not become a bottleneck. We have proposed a new consensus algorithm of the number map result, which makes the transaction number of the block very high, because Our block speed can reach four blocks within a safe range.
thank you all!
thank you all!
