With the launch of the second phase of the Filecoin testnet, the market popularity of Filecoin continues to rise, and various groups such as the technology circle, currency circle, mining circle, investment circle, and media circle are all rushing in. Miners from all walks of life are also gearing up to compete for a prize of up to 4.1 million FIL in the testnet incentive competition that will start on August 25.
As the launch date of Filecoin mainnet approaches, ordinary investors are paying more attention to Filecoin from the concept hype and promotion model to the comparison of the company's technical strength and various technical indicators. However, faced with a series of parameters such as effective computing power, block rate, packaging speed, computing power increment, luck value, failure rate, success rate, etc., ordinary investors are at a loss when faced with the propaganda of various manufacturers, and everyone claims to be the first. They don't know who to believe. In fact, for ordinary investors, the evaluation standard is very simple, that is, the return on investment. How to evaluate the return on investment? First, it depends on how many coins this mining machine manufacturer can mine; second, it depends on how many people the coins mined by this mining machine manufacturer are distributed to. Investors can calculate how many coins they can get based on their investment shares, and then calculate their return on investment based on the real-time coin price.What factors affect the coin production of mining machine manufacturers? Filecoin is storage mining. Miners obtain effective computing power based on how much data they actually store and submit proof of replication to the chain. The higher the effective computing power, the greater the probability that the miner will receive block rewards. Therefore, the core indicator for evaluating the technical strength of a miner is effective computing power. The faster the miner's packaging speed, or the more sectors that are packaged concurrently, the more effective computing power they can obtain per unit time. In other words, the faster the computing power growth rate, the greater the effective computing power will be. According to Filecoin's mining mechanism, the miners with greater computing power have the greater right or probability to obtain block packaging, so there are parameters for the vote winning rate and block generation rate. The vote winning rate is the probability of winning a vote, and the block generation rate is the probability of obtaining a block reward. This is also basically a positive correlation. The greater the effective computing power, the higher the proportion of the total network computing power, the greater the vote winning rate; the greater the vote winning rate, the higher the block generation rate, and the more block rewards are obtained. Some people here are wondering, why sometimes the node ranked first in effective computing power does not have the highest block production rate? Because obtaining effective computing power, that is, the process of generating replication proof to obtain computing power, and submitting time-space proof to participate in block packaging are two independent processes. When the effective computing power of miners reaches a certain level, the block generation will be relatively stable. However, the effective computing power ratio and the block generation rate are not absolutely corresponding. Since it is still in the testing stage, miners can only encapsulate sectors to increase computing power within a period of time without participating in block packaging, or because of problems in the process of submitting time-space proofs, the block generation rate is low. We can see from the block browser that when the time is long enough, the block reward of a miner is basically equal to the proportion of its effective computing power, that is, the higher the computing power, the higher the mining income. Therefore, for ordinary investors, they only need to pay attention to the two data of effective computing power and block reward, which basically represents the mining capacity of a miner. The second thing to look at is the mining efficiency of the miner, that is, how many mining machines the miner uses to mine these coins, and then how many people these coins are distributed to. Some mining machine manufacturers promote the packaging efficiency or mining efficiency of a single mining machine, which is actually misleading investors. Because Filecoin mining requires a cluster solution, a single mining machine basically cannot mine coins. First, according to the latest mechanism officially released by Filecoin, after the mainnet is launched, mining machines with effective computing power below 100TB will not have the opportunity to participate in block packaging, which means that they can basically mine no coins. To obtain 100TB of effective computing power within a reasonable time, the investment in hardware equipment alone is almost one million yuan. Therefore, the efficiency of single-machine mining is a false proposition. Secondly, a mining company can use dozens or even a few devices to participate in testing, but the difficulty and complexity of testing small clusters and large clusters are completely different levels and cannot be multiplied by simple coefficients. Recently, some block browsers have added an indicator of "24-hour mining efficiency", which is to calculate the mining income of a single T computing power by dividing the total mining income of a miner by his effective computing power. This algorithm is actually wrong. For example, in the above figure, we can see that among the top five nodes in terms of effective computing power, the mining efficiency of the first-ranked node "t01598" is the lowest, only 2.27FIL/TB, and the mining efficiency of the nodes ranked 2-5 is not much different; while the mining efficiency of the nodes ranked 162 and 163 is as high as 24.31FIL/TB and 27.72FIL/TB. This is obviously unreasonable.
Because the top-ranked node "t01598" was started late, it started to explode on the night of August 22, and its computing power increased rapidly, reaching the top of the effective computing power ranking in less than half a day. However, the process of miners obtaining effective computing power and participating in block packaging is two different processes, and there is a certain time difference in between. The effective computing power of the "t01598" node is only 736GB, and it has only received two block rewards; the "t0260320" node has only received one block reward. Due to its low effective computing power, it ranks first in the entire network in terms of so-called mining efficiency. Another situation is when the node loses computing power. For example, the mining income of the "t01657" node is as high as 268.18FIL/T. In fact, it is because the node was punished with 178TiB of computing power today . The numerator remains unchanged (the amount of coins produced per day), and the denominator suddenly becomes smaller (effective computing power), so the mining efficiency is very high. Therefore, it is an unreasonable calculation method to divide the miner's daily coin production by the effective computing power. What ordinary investors really care about is their return on investment, that is, how much money they invested in purchasing mining machines or cloud computing power, and how many coins they can get in a day. The basis for dividing coins should not be the effective computing power of the miner, but the total scale of the miner, which can be the number of mining machines or cloud computing power (total storage space capacity) of the company. Investors will be allocated profits in proportion to the proportion of the number of mining machines or cloud computing power they purchased to the total number of the company. For example, the computing power growth rate of Company A and Company B is 200TB per day, and they can mine 20,000 FIL every day. The number of mining machines that Company A is involved in is 500 (or 100,000TB of storage space), and the number of mining machines that Company B is involved in is 1,000 (or 200,000TB of storage space). Then the mining income that can be distributed to investors of Company A for each mining machine is 40 (or 0.2/TB); while the mining income that can be distributed to investors of Company B for each mining machine is 20 (or 0.1/TB). The return on investment of customers of Company A is twice that of customers of Company B. This again involves the number of mining machines of various mining machine manufacturers. Some block browsers have added a new parameter for mining machine equivalent , which is to regard the official recommended configuration of Filecoin (AMD Ryzen Threadripper 3970X, NVidia GTX 2080Ti, 128GB 2133mhz) as one benchmark mining machine, and the mining machine equivalent is the number of standard mining machines converted by the miner (the ratio of the miner's computing power growth rate to the growth rate of the benchmark mining machine). Some people use this mining machine equivalent as a reference value for the number of mining machines of a mining machine manufacturer, which is another wrong algorithm. Because different mining machine manufacturers use different hardware configurations and cluster solutions, and have different algorithm optimization capabilities, the packaging efficiency of each company's single machine varies greatly. For example, some companies can complete the packaging of a sector in 3 hours, while some companies can concurrently process more than 10 sectors. Therefore, we cannot simply use the official configuration and standard cluster method to infer the number of mining machines actually used by each mining machine manufacturer. Moreover, since Filecoin mining basically requires cluster solutions, most mining machine manufacturers use heterogeneous clusters, that is, separate computing and storage. The mining machine equivalent refers to the number of computing servers, not storage servers. The base for ordinary investors to participate in profit distribution should be storage space, that is, the number of storage mining machines or the capacity of total storage space. Among various indicator parameters, effective computing power is the core, which is the basis for obtaining mining income. Without effective computing power or with low effective computing power, you are not even qualified to participate in mining. Miners with high effective computing power have relatively stronger mining capabilities and higher mining income. Secondly, it depends on the scale of each miner's mining farm. The bigger the better, because the larger the scale, the larger the base of participation in the distribution, and the income that can be distributed per TB of cloud computing power (storage space) will be relatively smaller. Therefore, as an ordinary investor, on the one hand, you need to learn to view the mining capacity of each miner through a block browser, that is, the daily coin production; on the other hand, you need to pay attention to the sales scale of the miner and whether the mining machines (cloud computing power) it sells have corresponding physical mining machines delivered to participate in mining.
Currently, only the " CDSI Alliance " is known to have announced the total size of its mining pool as of the end of July. As of July 25, 2020, the total storage scale of the "CDSI Alliance" nodes is: 270,720 TB. Among them: the total storage scale of Space-Time Cloud Technology is 130,560 TB; the total storage scale of Lingdong Computer is 140,160 TB. During the testnet reward competition, the "CDSI Alliance" will use the above cloud computing power scale (total storage capacity) as the base for the distribution of FIL rewards for the testnet reward competition . If the "CDSI Alliance" receives a total of 270,000 FIL rewards during the testnet reward competition , approximately 1 FIL can be allocated for each TB of cloud computing power (storage space). * The rewarded FIL will be distributed linearly 6 months after the mainnet is launched. The specific distribution rules are subject to the specific rules officially announced by Filecoin. After the testnet incentive competition is over, Spacetime Cloud will also invite institutional customers, agents and some customer representatives to the computer room to conduct equipment delivery and acceptance work, to ensure that every mining machine we sell has a corresponding device delivery for every 1TB of storage space, and all data is open and transparent. In addition, each customer's equipment will also have an independent equipment code and delivery list. The testnet incentive competition with a prize of up to 4.1 million FIL is about to start, and major mining machine manufacturers are actively preparing for the competition. After the three-week competition, and after about 3 weeks of preparation before the mainnet launch, Filecoin's mainnet will be launched. Let us look forward to the beginning of a new era together! |
