As we all know, the payback period for ETH mining has exceeded 1 year... As the entire network knows, hard drive prices have experienced a roller coaster ride in just one week. Regarding storage coins, whether they can actually be stored is not discussed (or I don’t know enough to write about it) BTW, I don’t recommend any projects, I just want to see how to invest in the current popular storage sector projects at low cost/first-line Recently, many PoC/PoST projects have been promoted. Let’s first classify these coins into categories: Filecoin, Filecash, Filestar Arweave, Siacoin, Storj, Chia Crust, Phala Swarm BTT
Let’s define it first. Storage coin mining is based on a certain algorithm. Filling up the hard drive proves the “computing power”, and the unit is generally capacity. The process of filling up the hard drive can be simply understood as writing lottery numbers on the hard drive and then drawing a lottery (mining) This should be the most familiar to everyone ~ These should all be considered as incentive layers based on the IPFS network. Therefore: If you have sufficient funds and are familiar with the Filecoin economic model, it is recommended to buy mining machines from miners with independent development capabilities; if you are a retail investor, it is recommended to consider the platform endorsement and directly buy packaged computing power. (Because to be a qualified Filecoin miner, you must not only be able to install machines, systems, and deployment, but also maintain distributed storage systems. You even need to solve bugs in the official code and understand Filecoin's economic model to choose a better mortgage strategy. (There must also be enough funds to meet the pledge and gas fees to keep up with the growth of the entire network's computing power~) But! With the rise in the price of FIL, the current price of FIL single TB full storage computing power has exceeded 10,000 (I remember that the cloud computing power of various platforms was counted at that time (the price during the packaging period was basically around 2K/T. Later, the market price of full storage computing power was about 4k/T). If you cannot afford to invest, it is unnecessary. It is better to buy coins directly. Many platforms have also launched Fil financial management. The annualized rate is about 10-30%. Relatively speaking, among the forked coins, Filecash/Filecstar has a lower entry cost at this stage: Filecash supports both AMD and INTEL and is friendly to the existing market; the staking cost is lower, the gas fee is negligible, and the income will be relatively higher; but the original SDR algorithm has been changed, and entering the network requires re-packaging of computing power. Filestar has the same underlying structure as Filecoin and does not require repackaging. The computing power can be switched directly. There are already machines on the market that can perform dual mining. I don’t know much about mining, but I feel that the price fluctuations of the currency are quite like Mario. Mining requirements (official): AMD < 8-core CPU (with SHA), 128GB of RAM, powerful GPU (currently generally using 2080Ti/3080/3090 graphics cards), < 512GB SSD (1TB SSD NMVe interface) About sealing/full storage: Simply put, store data (PoRep), seal each sector when it is full, pledge the corresponding tokens (pre-pledge), and have the network administrator regularly check whether the stored data is complete (Post). At this time, quickly SNACKs calculations within 30 seconds to prove that the data is still intact (election-post), and then wait for customers to pick it up at any time until the order is completed and all the pledged coins are obtained. In addition to these, there is also a certain amount of gas fee to ensure smooth operation throughout the process, so as to avoid being penalized for arrears that affect network speed; as well as labor costs for software and hardware operation and maintenance. Features of this type of project:
Old brand projects around 2017/invested by top VCs/listed or about to be listed in big firms/overseas background/claimed to revolutionize storage/... If you ask, I don't know what they do. Anyway, the storage sector is rising wildly. If you ask, they are a star team and a well-known VC, going through bull and bear markets. After all, the breakthrough of the omnipotent blockchain technology is to break through the skylight of "computing and storage"! ! ! For technical analysis of these currencies, please directly link to this tweet: Which is the better decentralized cloud storage? Filecoin vs Storj design points and actual test PK Let’s get to the point, mining: Blake2b algorithm, now there are two mining pools (F2Pool & DXPool) and the following mining machines can mine 1. Obelisk SC1/SC1 Immersion Miner 2. Jinbei HS3/HS3 SE/HS5 SC&HNS dual mining machine The tutorial can be viewed in DXPool/F2Pool. It is said that HNS mining machines are very difficult to buy now. HS5 has a 50% premium. According to oral data, 60% of the orders are from overseas. All HS5 & CK5 intention orders will be stopped before July (it is all due to the shortage of chips...) (Obelisk's official website is pretty cool~) Decentralized cloud storage platform. ERC20. The second storage project of CB after FIL on March 25. The idea is a cheaper AWS S3 with faster uplink and downlink speeds, similar to Uber’s integration of idle resources for output. (Q: But idle space is always unstable..) Little-known fact: Storj reduces the time it takes to synchronize a new full node of ETC from 12 hours to 30 minutes, which is equivalent to building a full-node CDN. Mining requirements: minimum 1 CPU per node, 500G hard disk, 2TB/month bandwidth, at least 5MB/S upstream, at least 25MB/S downstream. Minimum cost is 800 yuan; 95% online rate must be guaranteed; bandwidth and computing power will be weighted to a certain extent. Post-mortgage (i.e. part of the income is pledged) is required, and all will be returned in at least 15 months. 1. Storj content storage fee is $10/TiB/month, each TiB has 2.7 times redundancy and is divided into 80 shares, and the miner reward is $1.5/TiB 2. The upstream and downstream traffic fee is $45/TiB, and the miner reward is $20/TiB. Ah~ here comes my favorite project (so I wonder if this article is subjective or not...) Keywords: the future Library of Alexandria; one-time payment, permanent storage; the storage protocol with the most cooperative NFT projects; Little-known fact: Recently, a payment platform based on Arweave has emerged. The development team is also very good. Currently, it only supports Ether. Ar's mining mechanism - RandomX+PoA, motivates miners to synchronize as much and as rare block data as possible. The hard fork on 2020/12/19 at the end of last year changed the original strong and strong CPU status. Let's explain the current mining logic 1. Perform a hash calculation (consumes CPU performance) 2. Random read data access verification in the hard disk (testing data storage capacity and hard disk read performance) 3. Perform a second hash calculation (consumes CPU performance) Therefore, AR mining requires a high-performance CPU + a high-performance SSD to support it. In fact, AR mining was basically unprofitable last year, but as the storage sector became increasingly popular, AR's coin price trended upward (0.5U - 29U), and AR miners once pushed up the SSD market price... Recommended configuration: AMD series CPU, solid state nvme SSD (the faster the random 4K read capability, the better) For example, AMD R9 3900X SSD: Samsung nvme SSD However, the current threshold for participating in Arweave mining in the country is very high. The main reasons are as follows: 1. Data synchronization is slow; 2. Data desensitization (force majeure) However, due to policy reasons, if domestic miners want to participate in AR mining without risk, they need to desensitize and encrypt the data synchronously and read as much data as possible. At present, the data is mainly encrypted and desensitized in China through third parties; 4. Malicious competition caused by profit-seeking mentality may lead to unhealthy growth of network data volume, resulting in the actual computing power being swallowed up. AR mining, the theoretical payback period for a new machine is about 6 months (currently the payback period for ETH mining is > 220 days) But the dynamic cycle should be getting longer and longer. However, the residual value rate of AR mining machines is extremely high: the CPU is directly converted to Monero (most AR miners are Monero miners at present); because AR mining requires high-speed data reading, but no disk scanning, there is almost no loss to the hard disk, and the loss is about 30%. [Note: AR computing power unit is KH]
Finally arrived at Chia! Pure dry goods! Quick!! The person involved is now looking at a bunch of servers and crying. As a PoC novice, I have spent three days in the github document and know that I must do it myself! No matter how good I am, it is useless! A Plot takes at least 6 hours, and if the progress bar is less than 100%, it is useless! ! ! First of all!! Chia is not for storage!!! Chia just uses the PoST (Proof of Spacetime) consensus mechanism to maintain the computing power of the entire network. In simple terms, it is "green BTC". The mining logic is that "farmers" all over the world "sow" in their own hard drives and "harvest" according to the amount of sowing. With the expansion of the "Chia vegetable garden", more "farmers" are needed to "sow". XCH is not an incentive token of a "protocol", but a payment method, the so-called "real value coin". This is undoubtedly the hottest hard disk mining project recently. Storj on CB is not as popular as it. Various Chia mining derivatives (joint mining/full storage mining machines/non-physical computing power, etc.) have invaded my circle of friends, Weibo, and group advertisements. Because the story is still long and can be written slowly~ so the length here will be longer Currently, the market has very high expectations for the price of Chia. Whether it is early investors, old PoC miners, those who made money in the bull market and want to cash out, investors who did not catch the FIL train, or various miners, mining pools and exchanges, they have all started to engage in Chia mining. Trivia: Chia founder Bram Cohen announced his resignation as CEO and left the BT team three months after BitTorrent was acquired by Torn. The first edition of the Chia Network green paper was released in July 2019: https://www.chia.net/assets/ChiaGreenPaper.pdf Economic model: 21 million XCH pre-mined, 64 XCH/10 minutes, halved every 3 years; total circulation in 50 years: 47,910,720 XCH; no ICO, no total issuance; I want to emphasize here that anyone who currently says they have private equity is either stupid or bad (there are no such stupid people at the time of writing). Just look at the mining guide. The official statement is that there is no ICO, investors are in the form of equity, and tokens will be used as product SEC. Chia's fundraising situation: It raised US$3.4 million in the seed round of financing in 2019, and in August 2020, it received US$5 million in financing from institutions such as Slow, Collab Crypto, IDEO and Naval Ravikant. Chia’s technical highlights: BLS/VDF/Chialips/Timelords Here is a collection: https://github.com/Chia-Network/chia-blockchain/wiki/Consensus-Algorithm-Summary Therefore, compared with the current popular mining projects, Chia's main features are No front-end or back-end collateral is required, no penalty mechanism is required, and mining can be done as long as there is data, just like AR; Expectations are extremely high. Projects in the same period include Cosmos, Filecoin, and Polkadot, which have been in development for up to two years. Official announcement: It will be listed on the SEC in July, with an IPO price of 1 pound/XCH; The development lasted for two years and it was officially launched on March 23, 21. Within 6 weeks of the launch, the wallet transaction function was suspended. Mining costs are low. It takes up almost no resources except storage. The investment threshold is low, and small-scale deployment can support home mining.
Founder Bram Cohen once changed the Internet. The BT protocol was once the largest transmission protocol for Internet traffic. Chia mining investment risks: Project stability risk: reduced returns Market bubble liquidation risk: reduced returns Hard disk mining equipment depreciation: the residual value of second-hand hardware within one year is about 30%-50% (this is the normal price!!)
Pay attention, how to participate in mining and calculate the investment payback period step by step! ! Hard disk mining generally consists of a server, CPU+RAM+SSD+multiple mechanical hard disks. Most hard disk mining projects require you to plot the hard disk first to prove the computing power (some just use the token/machine serial number to prove it directly), so professional plotters appear (need high-quality CPU + 512G large memory + 2 or more large-capacity SSDs). Miner is purely for mining, and the CPU and RAM requirements are not that high, so you can just stack hard disks without thinking. Chia mining, for every 101G file of drawing, the amount of data that needs to be written into the cache is 300+G. In other words, a lot of temporary files need to be generated during the P disk process. These are necessary processes for calculating the final file, so a lot of time is wasted on these temporary files. The academic reason for such a design is that "real-time generation of Plot" has become impossible to have economic value. Then, it is well known that Chia P disk is extremely difficult. So the question is, how to improve the efficiency of P disk? High frequency CPU + high speed large capacity SSD Back to the mining input calculation, the mining cost of Chia is mainly as follows: 3) Hardware operation and maintenance costs 4) Mining software operation and maintenance costs First, the Chia mining scheme is divided into two types, homogeneous and heterogeneous (borrowing the term FIL mining machine haha..) 1. Heterogeneous solution: Miner machine + Plotter machine Plotter machine: high-performance CPU >512G memory 2*8t SSD/single unit; Storage machines: Purely stacking hard disks and servers, only the cost of HDDs needs to be considered. Advantages: Single P disk is more efficient and reduces storage machine costs; storage machines can be expanded infinitely Disadvantages: Too customized, not easy to cope with the risks and changes of mining; speeding up the P disk speed, the actual large amount of cost is quickly lost. At present, I have seen some teams doing joint mining or P-disk services using such solutions. Dedicated machines are used for specific purposes and profits can be calculated more easily. Currently, computing power is growing relatively fast, but the P-disk capacity has a bottleneck and risk control is not easy. 2. Homogeneous solution: Miner-Plotter all-in-one machine Since there is no separate P disk, a high-performance CPU and large-capacity RAM are required, and SSD is used as a cache disk [recommended 2288, yyds] Advantages: mining and p-disk at the same time; low difficulty in operation and maintenance, simple expansion; strong risk resistance, Disadvantages: High cost per unit, increased platform cost; fixed investment, low P disk efficiency leads to computing power growth that cannot keep up with heterogeneous architecture. In summary, based on the experience of the technical guy and the big brothers who have invested in mining, whether it is a small or large-scale investment, the homogeneous solution (all-in-one machine) is more recommended: 1. The investment risk coefficient is the lowest. Professional P disk machines have efficiency bottlenecks. Whether it is investment/external, it is not suitable for scale expansion/reduction. 2. Operation and maintenance is simpler, and P disk and mining can be monitored at the same time. 3. When the memory is full, the CPU/RAM can be replaced with low-cost versions to reduce costs. BTW, if it is an external sales team, and the scale is not large, you can use heterogeneous, fixed p-disk efficiency, and low storage machine cost; you can provide p-disk services separately to share sales risks. Ok, now that the mining plan has been determined, the next step is to calculate the mining cost and payback period based on the machine purchase configuration and theoretical computing power on the market. First, let's talk about the price expectation of XCH. Since Chia is not circulated in the secondary market, the average daily income in this article is calculated based on the average transaction price of various OTCs. At present, according to the official IPO price of 1 pound per coin, the expected price of projects with similar backgrounds in the current bull market is $10-20. Assuming that the initial price is around $10, the payback period should be around 3 months (plus the P-trading time). Chia's total network computing power is 404.06PiB (100+PiB when the mainnet is launched) Total circulation: 21,324,996 XCH (pre-mining is 21 million, 324,996 XCH has been produced) Number of unique addresses: 5794 (the curve is almost smooth, with new miners joining every day) The output per TiB has been halved in the past 30 days, from 0.080XCH/day to 0.023XCH/day when it was launched. The payback period table has been deleted. Different configurations, different channels, different systems, different efficiencies, different times, and different costs! Note: This article was updated on April 21. The data is real-time data. Please analyze it based on the actual data of the entire network when you read it. Chia Block Explorer https://www.chiaexplorer.com/ (Dynamic payback period prediction table, not for public ~) Summary of this category:
Siacion: I don’t want to pay attention to Sia’s incentive mechanism and team. If you have a ready-made machine, go and mine. Jinbei’s HNS dual mining is really good, but I can’t buy one. Storj: Commercial cloud storage. You only need a 1T hard drive (around 200 yuan) and a 250 yuan Raspberry Pi to mine at home. The incentive mechanism is traditional and the profit model is not objective. Arweave: CPU+SSD, the cluster solution is more cost-effective. China currently relies more on the synchronization data speed of the mining pool. Chia: It is the hottest, please be careful to identify manufacturers with premium risk. Anyway, to be a miner, you should dig wherever there is a lot of activity! Polkadot ecological storage has not yet been put on the main network. The currency price is objective, Polkadot is bullish. I don’t have time to study the currency, so I’ll just give you the data~ Hard disk mining machines require pledge. The pledged coin is 1cru/T. The pledged coins can be directly withdrawn at unlimited intervals and released directly in about 7 days. 24-slot 8T, 192T hard disk mining machine, daily output of about 2.5cru Mining depends on the number of CPU cores, and requires staking Each core base is pledged with 1620 PHA. The pledged coins have no time limit and can be directly withdrawn in about 7 days. Currently, a 10-core machine produces about 180-190 PHA per day. The son of the Ethereum Foundation. The web3.0 native service layer is part of Ethereum's vision for a decentralized network. It will be put on the mainnet in June. The token is BZZ, with an initial supply of 62.5M. It is mainly used for bandwidth and storage rewards. Swarm means beehive. My understanding is that Bee (miner) needs to interact with Queen Bee (master node) constantly to establish connection, contribute bandwidth uplink and downlink, storage space, and discover new Bee (newly joined miner). Mining requirements also include bandwidth, processor, and hard disk. (I have never mined before, nor have I reached the first level, so I will not go into detail.) The testnet airdrop started in January with 1 million tokens (seems to be ending soon) Airdrop details link: https://medium.com/ethereum-swarm/important-notice-regarding-the-1-million-bzz-airdrop-d6b52ddd1af3 I don’t know about Mr. Sun’s coin and I don’t want to know about it. It seems that Sun Ge acquired the BitTorrent project of Chia founder. Although a customer said yesterday that the price of BTT rose by 60 points a few days ago, TAT The current income consists of three parts. The first is leasing, which is a contract and is paid monthly. The second is airdrop rewards, which require a certain number of packages per day, more than 100M, and the Chinese user group is pitifully small. The third is the accelerated sharing of torrent files. Torrent files have not been seen in China yet. Friends who have read this are in love with it! ! This year, the storage sector has regained its vitality, and the popularity of public chains has slowly declined. We are waiting for the various storage projects to compete! PoC is the best in the world! ! The HDD is solid, but the coins are changing! ! ! |
