Analysis of the optimal mining method for Filecoin

Editor: Remember to pay attention

Source: Filecoin Community

The following article comes from: Daoshuo Blockchain

The Filecoin testnet will be launched on December 11, 2019, and the mainnet is expected to be officially launched around March 2020. Friends who have been paying attention to and interested in participating in Filecoin mining should have noticed that various mining machine manufacturers have begun to promote it on a large scale.

Once I talked with the person in charge of a mining machine manufacturer. He said that there are two or three hundred companies in the country that make IPFS mining machines. According to my preliminary understanding, there are at least dozens of them, conservatively speaking. For most people who do not understand the technology and the details involved, how to choose is a big problem.

After all, investors are investing real money, which involves both huge benefits and huge risks, so they must weigh the benefits and risks comprehensively before making investment decisions.

After two months of investigation and analysis, we also conducted exchanges and research on more than a dozen IPFS mining machine companies on the market. In this section, I will try to share with you what we know.

When we analyze the investment value of any project, we must first look at its prospects and potential. The same is true for IPFS+Filecoin. Regarding its investment value analysis, I have already conducted a detailed analysis with you before, so I will not repeat it here. Overall, I am optimistic that this project is a very promising project. Its scale effect and consensus are likely to enter the top ten of the global cryptocurrency market value.

So is it possible that the actual application and expectations of this project after it goes online are not as good as we imagined? Or the price of the coin performs poorly after the project goes online, or the efficiency of all mining machines in mining coins is not high? This possibility is completely possible.

However, since the project has not yet been launched, we cannot determine this risk through data analysis. This is the worst-case scenario that every investor who plans to invest in Filecoin mining must prepare for. It is a risk that every investor who participates in the investment will face together. It is a systemic risk of this project. It has nothing to do with which company's products investors choose. It is not a factor that investors should consider when choosing a specific product, but a factor that investors should consider before participating in this project.

Then, the next step in analyzing the project is to see where the key factors for the project to make a profit are, that is, where is the threshold of the project? Is it the market, marketing, technology or other aspects?

In the IPFS+Filecoin project, the key point is technology. IPFS+Filecoin belongs to a brand new field, and both its technical framework and technical elements are brand new attempts. Although the project's code and algorithm are open source, the key to each mining manufacturer is how to use the best method to run this algorithm and calculate the result.

This is very similar to Bitcoin mining.

The algorithm for Bitcoin mining is completely open, but how to run this algorithm most efficiently and mine Bitcoin as quickly as possible is the specialty of each mining machine company and the focus of competition in this industry.

In the current Bitcoin mining market, Bitmain is the strongest, but this does not mean that Bitmain has a permanent advantage. In the early years, the mining machine invented by the young genius "Baomao" was more dazzling than Bitmain. It once monopolized 40% of Bitcoin mining, but later due to mistakes in the process of technological iteration, Baomao's mining machine withdrew from the stage of history and was taken over by Bitmain.

Therefore, technological iteration and competition play a central role in the competition for Bitcoin mining machines.

The same is true for IPFS+Filecoin.

We believe that in such a highly competitive field, only focused investment can give us a head start in mining.

According to our research, many companies selling Filecoin mining machines only regard the Filecoin mining machine business as a non-core business. In our communication with them, they do not think that technology is a decisive factor, and it will be difficult for such companies to win in this fierce competition. Therefore, the mining machines sold by such companies should not be purchased.

Therefore, when investors choose Filecoin mining products or services, it is recommended to choose companies that focus on Filecoin mining and technology research and development.

Since technology plays such a critical role in Filecoin mining, what technical solutions have companies on the market adopted?

There are two main solutions, one is single mining machine mining, and the other is cloud computing mining. Let's look at these two technologies separately:

1) Single mining machine

Simply put, single-machine mining is similar to Bitcoin mining. A mining machine is a machine. Investors invest in a physical machine, which includes hard disk, memory, graphics card, CPU, etc.

This composition is very similar to that of an ordinary computer, but the hardware performance and configuration are significantly different from ordinary computers. Its hardware needs to meet the minimum computing performance required by the Filecoin algorithm, so its performance will be stronger than that of an ordinary home computer.

However, the Filecoin algorithm is significantly different from the Bitcoin mining algorithm in that it does not require as much hardware as Bitcoin mining. In Bitcoin mining, only special ASIC chips can achieve significant efficiency, while Filecoin mining machines do not require such special hardware.

We can also say that the hardware configurations provided by various Filecoin mining machine manufacturers are publicly available on the market, unlike Bitcoin mining machines, whose chips need to be specially customized and can only be purchased from individual manufacturers.

So I think hardware configuration is not the technical focus of Filecoin mining, but the technical focus of Filecoin mining lies in software, system and operation and maintenance. And the system and operation and maintenance are also largely controlled by modern integrated management software systems, so it is still software.

This is not only the focus of Filecoin mining, but also the key point to distinguish between single mining machines and cloud computing mining.

When customers buy a single mining machine, in most cases they will entrust the mining machine to the mining machine manufacturer for operation and maintenance and unified management. These systems that manage a large number of mining machines are called mining pools. Mining machine manufacturers will also develop their own mining pool software to manage and maintain these mining machines.

This is very similar to the concept of Bitcoin mining pool.

2) Cloud computing mining

I mentioned earlier that the focus of Filecoin mining is on software, systems, and operations. It is precisely because different companies have different understandings and design concepts of software, systems, and operations that another mining method has emerged: cloud computing mining.

This design idea is to consider the mining activity as a whole.

Although IPFS+Filecoin is a storage business, in this system, the mining efficiency and reward factors are determined not only by the size of the hard disk but also by a series of complex factors. So in this case, will it be better to take various factors into consideration and manage, allocate and optimize them in a unified manner?

So some companies adopted this solution.

In this solution, since the system is a whole, if investors visit the site, they may not see mining machines with uniform specifications, but a large system composed of various systems and modules.

This is somewhat similar to the difference between an army group (cloud computing power) and a square queue (single mining machine).

Another approach in cloud computing mining is to aggregate several individual mining machines together into a large computing pool. This is actually still a single mining machine mining model.

So basically these are the two technical architectures at present.

So which of these two methods has more technical advantages?

Before analyzing the advantages, let’s first take a look at how Filecoin is mined and how to get rewards. Only by understanding the principles of Filecoin mining can we see the key points and which technology is more advantageous.

In the IPFS+Filecoin system, when a user pays Filecoin to the system to request the data he wants, the request is submitted to the entire network, and then the miners of the entire network begin to compete. In this competition, miners have a key data to submit to the system, and this data is the "reproducible proof". The "reproducible proof" is a proof that a miner has indeed stored certain data.

Assuming that the three miners closest to the user submit the correct "reproducible proof" to the system at the same time, the miner who submits first and is recognized will be selected to serve the user and receive a reward.

Therefore, in this process, calculating "reproducible proof" is the key. Note: This "reproducible proof" is a calculation process, so what factors are related to the speed of this calculation process? CPU? Memory? Graphics card? Or is it related to the overall performance of the system? There is no absolute standard at present. This is what all mining manufacturers, whether cloud computing power or single mining machine companies, are competing and estimating.

In addition, how to submit the result as quickly as possible after calculating the "reproducible proof" is related to network performance and system performance.

Therefore, the calculation and submission process of "reproducible proof" involves multiple factors. There is no consensus on which factor is decisive. It is the focus of competition and trade secret among manufacturers.

In addition, the actual efficiency of mining coins is also involved in this process. We can understand efficiency in this way: suppose a large node (10T) with strong computing power mines 10 coins, while another small node (1T) with weak computing power mines 1 coin. On the surface, the large node mines 10 coins per point, while the other small node mines only 1 coin per point. The large node mines more coins than the small node, but when averaged out, 1T in the large node mines 1 coin, and 1T in the small node mines 1 coin, so the efficiency of the two is actually the same.

So in the competition of efficiency, are larger nodes with stronger performance more advantageous or are nodes of different performances similar? Currently, it is impossible to draw a definitive conclusion based on authoritative data.

We have analyzed these aspects, but unfortunately we have not had authoritative data to support them. In fact, this is not only embarrassing for us, but also the root of the chaos in the entire industry. Because there is no authoritative data, everyone has their own opinions.

So is there really no data to refer to? Yes, and currently there are only the Development Network and Lotus Network.

The IPFS+Filecoin project has developed a development network for various mining businesses to test. However, this development network is constantly changing, and the data will be cleared every once in a while. Although this is an informal testing platform, as businesses that really want to take root in this industry for a long time, we think it is impossible to ignore this testing platform, so we think the data on this testing platform is still of certain reference value. Join the IPFS/Filecoin community and contact VX: liandaoxixi

In fact, several Chinese mining companies have tested mining on it and have achieved very high rankings. Of course, this result will change every period, and the computing power ranking of mining nodes will change every period.

So is this ranking the only reference standard? I think it is definitely not. Because what we see on this platform are just individual nodes, and we have no idea what configurations and methods these nodes use to obtain their computing power. It is possible that some nodes will spend a lot of money to use extremely powerful equipment and systems to participate in the test in order to win a high ranking, but the price of this system is impossible for ordinary investors to accept, and it is impossible that it is the mining machine or computing power actually sold to customers.

Therefore, we look at the development network and Lotus network in combination. We believe that although their data has limited reference value, the strength of manufacturers who have never participated in the test ranking or have participated in the test but have never achieved a good ranking on it is questionable.

When I analyzed this project before, I analyzed it from a technical perspective, but in addition to technology, there is another point that investors must consider, and that is the investment cost.

The return must be good but the cost must be low. If the return is high but the cost is too high, then the overall situation may not be suitable for investors.

Currently, the investments in the two mining methods, single mining machines and cloud computing power, are different.

Generally speaking, single-unit mining machine manufacturers sell mining machines directly to customers, and the mining machines always belong to the customers, but after Filecoin goes online, customers have to pay operation and maintenance fees to the manufacturers. Cloud computing power manufacturers rent computing power to customers, usually in T units, with a price per T, and a lease period of one year. During the lease period, customers do not need to pay any other fees.

Generally speaking, from a cost perspective: customers who buy mining machines pay a large one-time amount, and the mining machines belong to the customers. They have to pay operation and maintenance fees every month, but the unit price is lower per 1T. Customers who buy computing power can purchase flexibly. If they only buy 1T at a time, the amount they pay is smaller, but the computing power is only for one year, and no other fees are paid, but the unit price is higher per 1T.

Since neither the Filecoin testnet nor the mainnet has been launched, there are too many uncertainties, which make it difficult to analyze the entire industry. However, there are some basic logics and rules that can still be used for reference.

Once the Filecoin mainnet is launched, if the project is generally positive, it will inevitably drive up the price of the coin, and the rise in the price of the coin will inevitably drive more competitors to enter the market or more miners to participate. In this case, the supply of coins is limited, but there are more and more participants, which will inevitably lead to more and more fierce competition in mining. Therefore, we judge that the mining bonus of Filecoin is in the early stage.

This is the same as Bitcoin mining.

Therefore, participants who can successfully mine in the initial period after the mainnet is launched will make the most profit.

There are too many uncertain factors that determine Filecoin mining. For mining companies, investing energy in research and development when many factors are uncertain before the main network is launched is like crossing a river by feeling the stones, which is a considerable risk. The risk of complete failure cannot be ruled out. Therefore, companies that are exploring this field are in great uncertainty.

However, I think that if a company does not actively participate in R&D at this time and hopes to wait until the test network or main network is launched before joining the battle, the success rate is minimal. Because the Filecoin algorithm has been made public, although many factors are still changing, if you do not participate in the testing of the development network during this process and do not explore the impact of various factors on mining efficiency, it will be difficult to win if you want to cram before the final exam.

Filecoin mining will be an extremely hot project next year, and it will also be a field with very fierce competition.

I am very cautious about any statement that one technical school is absolutely better than another, because we can only see which technical school is better after the main network is launched. Of course, readers can develop their own opinions and make their own judgments based on their own experiences and understanding of technology. I hope that each manufacturer will strive to make the best technology in their respective fields.

It doesn’t make much sense to say who is better or worse now. When the main network is officially launched, it will be clear who has higher mining efficiency, and then the market will automatically tilt the balance towards the party with real strength.

As for whether to choose a single mining machine or cloud computing power, I suggest that investors choose the method that suits them according to their actual situation, but at least choose the top companies with high rankings.

I would like to emphasize one more point here. There are many people in the market who are taking advantage of the fame of IPFS and using a pyramid scheme to sell so-called mining machines. I will share this separately later, and I hope everyone will remain highly vigilant about this.