leaf

#

(1) Review of Mining History

Since the birth of Bitcoin, BTC mining has gone through five stages: CPU mining, GPU mining, FPGA mining, ASIC mining, and large-scale cluster mining.

On January 3, 2009, Bitcoin's founder Satoshi Nakamoto mined the first genesis block using his computer's CPU. For about a year afterward, the BTC network relied on CPU mining to maintain normal operation. A typical Intel CPU mining speed is 20μ～40MHash/s [1]. Since CPUs are designed to handle complex general instructions (CISC), their computing speed for the SHA256 algorithm used in mining is not ideal.

1. 1. GPU Mining Era

Due to the architecture of GPUs, which utilize a large number of parallel processing cores, the processing speed for the simple SHA256 algorithm is relatively fast. In December 2009, the idea of GPU mining was first proposed on the official Bitcoin forum. On September 18, 2010, the first GPU mining software was released, marking the beginning of the GPU mining era. Due to architectural differences, AMD GPUs have a mining speed about five times that of nVidia. A commonly used AMD GPU has a mining speed of approximately 300μ～400MHash/s, and graphics cards suitable for mining are often sold out. However, the drawbacks of GPUs became increasingly severe, with a computer equipped with six GPUs consuming around 800W. Not only is it difficult to set up and deploy such computers, but they also generate significant heat, making large-scale deployment unsuitable for ordinary households.

1. 1. FPGA and GPU Coexistence Era

By the end of 2011, FPGA-based mining devices emerged. An FPGA chip has a mining speed of about 200MHash/s, with power consumption reduced to 1/40 of that of GPUs. However, the high cost and complexity of deployment meant that FPGAs became toys for a few knowledgeable miners. During this period, FPGAs and GPUs became the main forces in mining.

1. 1. ASIC Mining Era

History always moves forward. When the energy consumption ratio (performance/power) of GPU mining reached a bottleneck, higher-speed, lower-energy tools emerged—ASICs (Application-Specific Integrated Circuits). It is foreseeable that ASICs will be the ultimate equipment in this game, ending the dominance of GPUs in Bitcoin mining.

(2) Hash Rate Fluctuations and Mining Pool Shares

As the difficulty of Bitcoin mining increases at a record pace, the hash rate of the Bitcoin network has increased from 20THash/s to 7PHash/s over the past year, nearly a 350-fold increase. This is mainly due to existing manufacturers such as Avalon, ASICMiner (AM), Butterfly Labs (BFL), Bitfury, and KNC, which uses 28nm technology. Currently, this computational power has reached 300 times the total computational power of the world's top 500 supercomputers.

During this process, the centralized mining—mining pool model has received continuous attention: if a mining pool can occupy 51% of the total network hash rate, it can launch a 51% attack on the entire network, enabling double spending, making all transactions on the network untrustworthy, which would be a devastating blow to Bitcoin. Initially, ASICMiner mined in the BTC Guild pool, and due to its excessive hash rate, it directly led to BTC Guild's total hash rate approaching 51% of the entire network. To mitigate risks, BTC Guild raised fees to drive away some miners, including ASICMiner. Later, ASICMiner built its own pool and used BitMinter to decentralize itself. So far, GHash, with a hash rate of 2.2PHash/s, occupies 31% of the total network, surpassing BTC Guild to become the largest mining pool globally. 50BTC suffered a hacker attack, leading to a sharp decline in hash rate share. Deepbit's share dropped from 20% of the network a year ago and 30% 18 months ago to 0.29% due to outdated mining protocols and distribution models. P2Pool's share also fell to about 1.7%. Additionally, the largest domestic mining pool, F2Pool, ranks among the top six in the network.

The exponential growth of network hash power has a significant impact on the speed of Bitcoin issuance. The continuous increase in hash power has made finding blocks more difficult, ultimately lowering the average block discovery time. Since new bitcoins are awarded for each block, the expected block discovery rate in the third quarter is about 7 minutes on average, rather than 10 minutes. If the network growth is not too exaggerated, more new bitcoins should have been discovered. This trend is likely to continue, as more mining devices are expected to impact the network.

(3) Mining Machine Companies and Development

ASICMiner

ASICMiner is a company virtually listed in the Bitcoin industry, headquartered in Shenzhen, China. The company has a total of 400,000 shares, with 163,962 shares circulating in the trading market. Shareholders enjoy 100% of the profits after deducting company operating costs and investment products, with profits sent directly to shareholders' linked Bitcoin addresses, distributed weekly.

ASICMiner has produced devices with a hash rate of about 200THash/s. Most sales are through third-party distributors, such as Block Erupter Blades operating at 10 GHash/s and USB devices operating at 300 MHash/s. On February 14, 2013, ASICMiner had a hash rate of 2THash/s, and from that moment on, it continued to increase its hash rate to 40THash/s. Although the overall network rate has significantly increased in the past few months, ASICMiner has also kept pace, striving to maintain 20% of the hash rate (approximately 720BTC/day). Their weekly dividends are about 0.013BTC/share, and since February 28, their total dividend per share has been about 0.386BTC, while ASICMiner's stock price once exceeded 4BTC/share. Additionally, ASICMiner has launched a franchise mining farm plan, allowing major shareholders to deploy machines for mining by using pledged shares. However, as several competitors began large-scale shipments and lower-process challengers emerged, the overall network hash rate surged, and ASICMiner's share continued to decline, currently accounting for about 1% of the total network, with a stock price around 0.3BTC/share.

Avalon

Avalon initially entered the market to prevent BFL from becoming the sole ASIC supplier in the industry. In September 2012, Avalon announced that it had begun producing 110nm chips, with a single chip speed of 280Mhash/s and a single machine (3 modules) speed of 60Ghash/s for Bitcoin mining. The first batch of ASIC miners was sent to customers in February 2013. The entire industry was excited.

Avalon began its first pre-sale on September 23, 2012, priced at $1299, with a total of 300 units. These machines began to arrive around March 3, 2013, and miners could recover their costs in just a few days. At that time, Avalon’s market price was inflated to 10 to 20 times the official price.

The second batch of 600 Avalons was pre-sold at a price of $1499 (equivalent to 75BTC/unit), with domestic support for payment in RMB, priced at 9300 yuan/unit. According to the agreement at that time, the second batch was supposed to be shipped in early April, but the actual shipping time was delayed until early May to mid-June. Even so, the second batch of Avalon customers enjoyed substantial profits—they could recover their fiat costs (i.e., $1499/unit or 9300 RMB/unit) in about 5 days or recover their Bitcoin costs (75 BTC/unit) within 15 days.

The third batch of 600 Avalons was still pre-sold at 75BTC/unit, but by the time of the third batch pre-sale, the price of BTC had risen to $80, making the price relatively expensive from a fiat perspective. The agreed delivery time was early May, but it was actually delayed until late July. During this period, the overall Bitcoin network hash rate grew so rapidly that customers receiving the third batch of Avalons would take about six months to mine back 75BTC, significantly reducing the return compared to the first and second batches.

Avalon also conducted bulk sales of ASIC chips, with each chip defaulting to a speed of about 282MHash/S, starting at 10,000 pieces, priced at 780BTC. The chips were opened for reservation in mid-April, with a delivery date of about 9 to 10 weeks later. The reservation volume reached over 900,000 pieces. However, due to packaging, testing, and customs issues, there were several weeks of delays, coinciding with the period of exponential network growth, which frustrated their customer base. Avalon offered a full refund option for customers who still had not received their pre-ordered chips. TerraHash (Avalon's distributor) had over $500,000 in orders, but the company decided to dissolve and refund 50% of the pre-order prices to customers.

Avalon's second-generation chips use 55nm technology, with a single chip speed of 1.6GHash/s, and they have begun selling chips and machines. These chips are fully compatible with the first-generation PCB boards, allowing customers to wait for their boards without receiving their first-generation chips and easily upgrade.

Looking ahead, Avalon is producing a 2U standard server based on the second-generation chips, which means ASIC miners can be deployed on a large scale in existing IDC data centers. Additionally, Avalon is developing third-generation chips.

Butterfly Lab (BFL)

BFL began pre-selling ASICs on June 23, 2012. They decided to use 65nm technology to compete against ASICMiner and Avalon, both of which used 110nm technology. Theoretically, 65nm can provide faster speeds and higher energy efficiency, but the design and tape-out process are more complex and require higher costs.

Ultimately, BFL delayed shipping the first mining machine to users for 10 months, during which they continuously told customers to wait just two weeks for delivery. This constant delay can be seen as a form of fraud. In early June 2013, BFL finally shipped a small batch of 5GHash/s machines, but these orders were effective as of June 5, 2012, resulting in a full year of delay. Despite announcing the complete end of claims that had accumulated until September 30, 2013, BFL still had a large number of customers waiting for the delivery of 65nm ASIC miners. Refund activities organized by several community members, through PayPal and credit cards, generated over $330,000 in refunds. Although their first-generation miners had production issues, BFL announced in mid-August that they would produce 28nm ASIC miners. They predicted the initial delivery date would be in December, and any existing orders were not expected to arrive before February. Given the production delays of their 65nm products, these predicted dates should be viewed as delays for a large amount of margin.

Bitfury

Bitfury uses 55nm technology, with a single chip speed of 5GHash/s, and was the first in the industry to reduce the power consumption ratio below 1W/GHash. Bitfury adopts a business strategy of selling chips and finished products. Bitfury began shipping its first batch of 55nm miners in early September 2013. They produced chips with a hash rate of about 500 THash/s, using 200 THash/s internally. The remaining approximately 100 THash/s will be sold to individual customers through retailers in the US and EU, while 200 THash/s was sold to a publicly listed mining company, 100TH. 100TH did a lot of work developing a motherboard chip that could use the 100 THash/s retail product. Since the successful chip launch in June, Bitfury's private pool ghash.io has seen its share of the total network hash rate continuously increase. Currently, the hash rate is around 2.2PHash/s, accounting for about 31% of the total network hash rate, surpassing BTC Guild to become the largest mining pool globally.

Active Mining

Active Mining is a publicly listed Bitcoin ASIC mining equipment manufacturer. In June 2013, they announced a collaboration with an unnamed semiconductor company to sell devices using various chip sizes. In September, they will only retain the 28nm product line. It has been revealed that they have teamed up with eASIC, a respected semiconductor company.

They expect initial shipments to begin in November, and any products ordered now are expected to be delivered in December. Active Mining has surprisingly low pre-sales, totaling less than $1 million, significantly below the production cost of a batch of 28nm ASIC chips. If Active Mining's plan proceeds, facing potentially massive production numbers, they may have a large inventory of devices to sell.

1. 1. CoinTerra

Ravi Iyengar, a CPU architect at Samsung Electronics, first heard about Bitcoin in May 2013. He then quit his job and raised $1.5 million to establish the mining hardware manufacturing company CoinTerra, which has since pre-sold hardware worth over $5 million.

He stated that the intensity of the arms race in the semiconductor industry cannot keep pace with Bitcoin mining; one month of Bitcoin mining is equivalent to a year in the semiconductor industry. Miners now realize that the real money is not made by them but by mining machine manufacturers. CoinTerra believes that in the next three years, the budget for investment in Bitcoin mining chips will easily exceed $100 million annually. Currently, the Bitcoin network has mined 11 million BTC, worth about $2.4 billion at recent values. In the next three years, the Bitcoin network is expected to mine 1.4 million BTC annually, which, at current values, is equivalent to $280 million per year.

The Bitcoin industry can be divided into upstream, midstream, and downstream

Observing the global development of Bitcoin, the entire Bitcoin industry can be divided into upstream, midstream, and downstream. The upstream refers to Bitcoin production. Most businesses surrounding Bitcoin mining belong to the upstream sector, such as the design and production of mining machines, the construction of mining farms and power generation equipment, and the operation of mining pools. The midstream of Bitcoin involves trading and storage. The downstream refers to the applications of Bitcoin.

There is also another perspective that places Bitcoin storage in the upstream sector because the storage issue needed to be addressed from the moment Bitcoin was born. So, this perspective is also valid. Our definition today places Bitcoin production in the upstream.

Mining is a topic we will discuss. In this lesson, we will talk about mining machines and miners. Through this lesson, you will learn how to select mining machines and understand the conditions required to be a good miner.

When Satoshi Nakamoto invented Bitcoin, he hoped that everyone could participate and did not want specialized mining machines to emerge. You may have heard stories about people mining at home or in the office with laptops during the early days, when CPU mining was still prevalent. Later, as computing power increased, some people came up with the idea of using GPUs for mining. Who was the first to mine with a GPU? Yes, it was the programmer who traded 10,000 bitcoins for pizza. At that time, GPU mining led to a global shortage of graphics cards, and this situation was repeated during Ethereum mining. In June 2017, as Ethereum's price rose, the graphics card market once again faced a supply shortage. I searched for graphics cards on JD.com, and it showed that all but a few high-end graphics cards were sold out, causing the stock price of chip company AMD to rise directly by 7%.

The Origin of Mining Machines

Returning to the early days of the Bitcoin industry, as the computing power required for mining continued to rise, GPUs reached their limits. To break through this limitation, specialized mining devices were invented. Although these devices are computers, they can only perform Bitcoin mining and hash calculations; we call them "mining machines." Bitcoin mining machines can only perform calculations for Bitcoin algorithms. Litecoin mining machines can only perform calculations for Litecoin algorithms; they are not interchangeable. It's not that it's technically impossible, but the essence of mining is a competition of calculations, a process of continuously breaking through computational limits. All conditions must be optimized to win in competition. If you try to balance everything, it will only reduce operational efficiency, and the computational power will not keep up with specialized mining machines, leading to market elimination.

How to Choose a Good Mining Machine

Currently, companies that produce mining machines are basically chip manufacturers. Mining machines consist of chips, cooling fans, batteries, and other components, among which the core component that determines whether a mining machine can mine more coins is the chip. The two most well-known Bitcoin mining machine manufacturers globally are Bitmain's Antminer and Zhang Nangen's Avalon miner. The latter is also the inventor of the world's first ASIC chip mining machine. ASIC refers to integrated circuits designed and manufactured to meet specific user requirements and the needs of specific electronic systems. There are also other well-known mining machines, such as the once-popular KNC Miner. This Swedish veteran Bitcoin mining machine manufacturer KNC Miner raised approximately $32 million in funding but unfortunately announced bankruptcy in May 2016. Additionally, other notable domestic mining machines include Long Mining Machines and Lightning Mining Machines. Because mining machine chips require very strong R&D technical capabilities, they need to race against the continuously rising global hash rate and keep up with technology. In the early days of the industry, many newly launched mining machines were eliminated by the market, and some were even phased out during the production process, resulting in significant losses for entrepreneurs and investors. Therefore, when choosing a mining machine, it is also essential to select strong brands and teams.

Given the intense competition in mining machines, how should one choose a mining machine if they want to mine? Theoretically, the best choice is to select the latest model. The latest mining machines tend to have lower power consumption and higher hash rates. When selecting a mining machine, consider the hash rate, power consumption, and historical reputation, including machine stability and after-sales service. The hash rate refers to a machine's computational ability, or how many hash calculations it can perform per second. Currently, mainstream Bitcoin mining machines have a hash rate of 14T, meaning they can perform 14*10^13 hash collisions per second.

Power consumption is an indicator of the amount of electricity a mining machine consumes during operation. Power consumption has a significant impact on mining costs. A mining machine typically operates 24 hours a day, so even a seemingly small difference in power consumption can lead to substantial cost differences over a year.

Here's a little knowledge: digital assets using the same algorithm can be mined with the same mining machine. For example, a Bitcoin (BTC) mining machine can also mine Bitcoin Cash (BCC), and an Ethereum (ETH) mining machine can also mine Ethereum Classic (ETC).

At this point, you might be wondering if it's feasible for me to mine Bitcoin myself. A few years ago, it was quite feasible. However, now it is very difficult for individuals to mine Bitcoin. This mode of individual mining is called the SOLO mode. Currently, mining is mostly done on a large scale, and cooperation with mining pools is required, so individual Bitcoin mining is more of a thought than a reality. I will explain the concept of mining pools later.

How to Participate in Mining#

If you can't mine yourself, you might want to know how to mine with a purchased mining machine. In fact, the process is simpler than you might think. Buy a mining machine, start it, and connect it to the network in three steps. First, choose a mining pool, register an account in the pool, fill in the sub-account, and set a password; second, fill in your Bitcoin receiving address. The bitcoins you earn will automatically be sent to this address. The address can be your wallet address or a deposit address on a trading platform, making it convenient for you to sell promptly and reduce transaction fees. Third, plug in the power and network cable to the mining machine, then turn it on to start operation. After powering on, the mining machine automatically assigns an IP address, scans the latest connected device IP addresses, and enters the backend. Then, set the mining pool account. After the setup is complete, it will start running in about half a minute, and you can see the real-time status of the mining machine on the backend page. The display of the mining machine is quite user-friendly, with basically two indicator lights: one shows normal operation, and the other indicates whether there is a fault. When the fault light comes on, you need to check if there are any issues with the mining machine's operation.

Let me share a little anecdote. A friend of mine loves to find good deals on mining machines on second-hand platforms. I asked him how he manages to find good deals on mining machines. He told me that many people, after learning about Bitcoin, like to buy a few mining machines online to try mining at home, thinking that even if residential electricity is a bit expensive, it doesn't matter if the cost recovery takes longer; after all, it's about participating in this experiment. However, once they plug in the mining machine, they find the noise is unbearable. So, within a day or two, either they or their family can't stand it anymore, and they are forced to sell the mining machine at a low price on the second-hand market. Usually, these mining machines are quite new and have hardly been used, making it a great deal to buy such second-hand machines. Every time my friend finds a good deal, he says that it's not the electricity costs that defeat home mining, but the noise.

Alright, let's review the main content of today. The development of mining machines has gone through three stages: from CPU to GPU, and then to specialized mining machines. I also introduced the tips for selecting mining machines, emphasizing three key points: first, look at the hash rate; second, consider power consumption; and third, check historical reputation.

If you want to mine, it's even simpler: choose a good mining pool, register an account, set up your wallet address, and plug in the power and network cable to start mining.

You will learn how to operate a mining farm, what is needed to operate a mining pool, and how to cooperate with mining pools.

Costs of Mining Farms#

The costs of a mining farm include construction costs, equipment costs, maintenance costs, network costs, and other costs. Maintenance costs include electricity and labor costs, which can vary significantly. This is why mining farms choose to build in areas with relatively cheap and stable electricity prices. Once a mining machine is running, it operates 24 hours a day, so having a team that can work in shifts 7*24 hours is also a necessary condition.

Now, with the professionalization of mining farms, how a mining farm is designed, including the advantages and disadvantages of water cooling, air cooling, and circuit design, greatly affects the maintenance, costs, and profits of the mining farm in the future. Early mining farm operations were relatively crude; they would just set up a frame and place the mining machines on it to start operations. Later, it was found that this operational method was not feasible, as the damage rate of mining machines was severe, and maintenance costs were too high. Ventilation and dust-proof solutions were introduced, and later developments included strict control of indoor temperature and humidity. The entire operational plan for mining farms is still continuously upgrading and evolving. Since mining machines generate a lot of noise, a single mining machine operating at one meter away can produce noise levels around 73dB(a). The impact of thousands of mining machines running can be imagined, so some places have been designed and renovated to create silent mining farms.

Operating a mining farm also carries risks, including policy risks, technical risks, and initial investments, all of which involve significant amounts. The payback period for a Bitcoin mining machine varies based on the price of the coin and changes in hash rate, generally ranging from 200 to 300 days. Coupled with maintenance costs, investing in a Bitcoin mining farm requires a strong ability to withstand risks.

Definition of Mining Pools

Having discussed mining farms, let's talk about mining pools. Compared to mining farms, you might be more curious about what a mining pool is. A mining pool is a collection of hash power. Initially, there was no role for mining pools in the Bitcoin world. Miners calculated on their own and packaged transaction information themselves. As more people joined mining and hash power increased, the probability of an individual miner successfully mining Bitcoin decreased, but more importantly, the income based on probability became increasingly unstable. An ordinary miner or mining farm might take a month to mine a single block. This poses a significant financial risk for miners. Thus, someone thought, why not pool everyone's hash power together? This would greatly increase the probability of mining a block, and then the block rewards could be distributed among everyone. Thus, mining pools were born. Mining machines are responsible for performing hash calculations, while mining pools are responsible for packaging transaction information. In other words, the actions of hash power competition and bookkeeping are carried out by mining machines and mining pools, respectively.

So how do mining pools convince miners to connect their mining machines' hash power? Here, I will introduce two mainstream cooperation models of mining pools: PPS and PPLNS. The PPS model (Pay-Per-Share) means that based on your hash power share in the mining pool, you receive a basic fixed income daily. Note that this is a fixed income. Suppose your hash power is 1T, and the total hash power of the mining pool is 100T, then you occupy 1% of the pool's hash power. The mining pool estimates that based on the current mining difficulty and the total global hash power, it can mine about 4 blocks a day. Note that this is an estimate, not a guarantee of 4 blocks. Each block reward can yield 12.5 bitcoins. Therefore, the mining pool can earn 50 bitcoins a day. The mining pool will pay you 50*1%, which is 0.5 bitcoins as your reward. Thus, even if the mining pool only mines 1 block today, you still receive 0.5 bitcoins, which is paid by the mining pool. However, if the mining pool is lucky and mines 10 blocks, you still only receive 0.5 bitcoins, meaning the mining pool profits more. However, in the long run, you and the mining pool are generally balanced.

The PPLNS model (Pay Per Last N Shares) means that "rewards are paid based on the last N shares." This means that once all miners mine a block, everyone will share the newly minted bitcoins based on their contributions, i.e., the proportion of their hash power. So how does it differ from the previously mentioned PPS model? In the PPS model, the Bitcoin dividends are paid in advance by the mining pool to miners, so as long as the miner's machine speed is stable, the number of bitcoins received daily will be very stable, which can be called a fixed investment-fixed return model. In the PPLNS model, it is a fixed investment-dynamic return model. The amount of Bitcoin earned in a fixed time has a certain element of luck. If the mining pool can mine many blocks in a day, miners will receive more bitcoins than theoretically expected. Conversely, if the mining pool mines fewer blocks than theoretically possible, miners will earn very little during that time. If the mining pool performs poorly and does not mine any blocks for a long time, miners may not receive any rewards during that period.

Among the top 10 mining pools globally, about seven are Chinese mining pools. It is worth noting that just because mining pools are established by Chinese individuals does not mean that all miners connecting to these pools are Chinese. Russia is also emerging as a new force in mining, and South Korea has recently shown great interest in mining. BitFury, a Russian company, was initially a mining machine manufacturer but has now transitioned to a mining pool, ranking within the top 10. Another Russian mining pool, Russian Miner Coin, announced plans to raise $100 million to compete with Chinese mining pools. This mining pool is reportedly owned by Dmitry Marinichev, an internet advisor to Russian President Vladimir Putin. Additionally, North Korea, which has never had Bitcoin nodes, began to establish Bitcoin network nodes in May 2017. If you look at the global distribution of Bitcoin nodes on a Bitcoin explorer, you will find that the distribution of Bitcoin full nodes is very decentralized, with nodes shining across South Africa, Japan, North America, South America, and Africa.

Returning to our topic, mining. If you want to be a miner maintaining a mining farm, it is relatively simple. Basic mechanical maintenance skills and the ability to understand circuit boards are sufficient. The life of a miner is quite simple, as they usually work in relatively remote areas. Therefore, those who can endure solitude and are technically inclined may be more suitable for this job.

Alright, in the sixth and seventh lessons, I have explained what the upstream of the Bitcoin industry looks like, including mining machines, miners, mining farms, and mining pools. You should now have a general understanding. Let's summarize:

Mining farms are gradually professionalizing, and the payback period is relatively long, requiring a certain ability to withstand risks. Thus, pooling everyone's hash power led to the creation of mining pools;

The main cooperation models of mining pools include two types: PPS and PPLNS. PPS is a fixed investment-fixed return model, where the mining pool pays miners in advance; PPLNS is a fixed investment-dynamic return model, where income is dynamically obtained based on each person's hash power share and luck;

The upstream of other digital assets is similar to Bitcoin; some have just introduced mining machines, while others have just established mining pools, and they are all still in the development stage.

1. 1. HashFast

HashFast collaborated with Uniquify and announced their 28nm chips in August 2013. They sold out their first batch of 550 units, estimating a hash rate of 220 THash/s and $3 million in sales. HashFast announced the start of testing for the 28nm Bitcoin mining chip on December 5, but some anomalies occurred. On December 13, during the third test, the chip passed testing, achieving a single chip speed of 500 GHash/S. HashFast expects the first batch of chips to leave the factory on December 17, with packaging starting on December 21, and they will soon be mailed to users.

1. 1. KNC

KNC was the first manufacturer to showcase 28nm ASIC chips. They originally expected to deliver in early September 2013 but showcased their single machine running at 500GHash/s on the evening of September 30. To date, KNC has shipped approximately 550 THash/s of mining machines. Meanwhile, KNC announced the next generation of 20nm ASIC solutions, pre-selling 3THash/s "Neptune" ASICs, which sold out within 24 hours for a total value of $8 million. KNC is currently collaborating with World Semiconductor to develop the next generation of 20nm/16nm products, expected to begin shipping in the first or second quarter of 2014.

1. 1. Grayscale Technology

In November 2013, Grayscale Technology, a domestic company specializing in the research and sales of Bitcoin-specific chips and complete products, officially signed a contract to settle in the Tianjin Economic and Technological Development Zone. The company will utilize 300,000 self-developed core chips to deploy the world's largest single Bitcoin computing cluster, controlling 10% to 15% of Bitcoin output globally.

1. 1. Silverfish Mining

Silverfish Mining was established in December 2013 as the world's first mining machine manufacturer specifically targeting Litecoin (LTC) mining. The first generation of Silverfish mining machines is expected to be released in the second quarter of 2014. Silverfish Mining implements equity distribution through Bitcoin block protocols, attempting to prove the feasibility of digital smart assets using the Bitcoin blockchain for the first time, giving digital smart assets real meaning and value. This attempt is likely to become the standard for the entire industry in the future, holding symbolic significance for the world.

(4) Outlook for the Mining Industry

For BTC, the proliferation of ASIC R&D companies and the launch of numerous ASIC chip mining machines, along with rapidly increasing difficulty, have driven up the overall network hash rate, mitigating the 51% risk and reducing the likelihood of a BTC crash. ASICs are a boon for Bitcoin. The price of BTC has also increased about 100 times from last year's 70 yuan (based on the average price in 2013).

The exponential growth of hash power in the network is unlikely to decrease before reaching 10THash/s by the end of 2013, and miners' investment return expectations will significantly decrease.

As customers begin to favor existing inventory products, pre-ordering will become less common. Compared to previous predictions of a three-month return, miners will be able to accurately calculate expected return rates without being overly optimistic.

In the long run, privately operated mining businesses can leverage economies of scale, which will reduce returns for individual miners.

Outlook for Bitcoin Development#

In the future, three macro trends are expected to have a significant impact on the degree of recognition of Bitcoin:

Speculative behavior in developed markets.

Capital controls and risks of collateral assets in the Eurozone.

Inflation in emerging markets, especially as the Federal Reserve is expected to reduce asset purchases and the labor market employment rate reaches a 30-year low, with stock and fixed income markets preparing for turmoil, which will positively impact alternative asset classes. Market participants will diversify, possibly by abandoning some traditional assets and incorporating Bitcoin into a small portion of their overall investment portfolio, and some institutional investors and advisors have already done so. New exchanges and investment vehicles, such as SecondMarket's Bitcoin investment trust, will make it easier than ever.

As controls on capital flows increase, governments around the world are looking to prevent shaky economies and currency devaluation from putting citizens' personal property at risk through restrictions. The most tangible example occurred during the Cyprus financial crisis when bank deposits were frozen and ultimately reduced by 48% to support bank capital adequacy ratios.

Similar situations have begun to appear in other parts of Europe, with Poland recently confiscating half of its citizens' private retirement funds to make its debt ratio more favorable. The Italian government crisis has also put its oldest bank in need of rescue. Even China's economic data shows signs of weakness and has implemented strict capital controls.

2011 Market

On January 27, 2011, the largest Bitcoin transaction occurred: three bills from Zimbabwe exchanged 4 BTC for 1 trillion Zimbabwean dollars in an over-the-counter Bitcoin transaction.

On February 9, 2011, the BTC price first reached $1, equivalent to the US dollar. The news that BTC was equivalent to the US dollar received extensive media coverage, leading to a surge in new users. Within the next two months, Bitcoin trading platforms for exchanges with the British pound, Brazilian real, and Polish zloty opened one after another.

On March 1, 2011, after four months of negotiations, Mt.Gox was transferred to its current investors, and after another four months of improvements, its trading volume surged, and prices skyrocketed, reaching a market share of 76%.

On March 6, 2011, the total network computing speed reached 900GHash/s but quickly fell by 40%, and GPU mining began to gain popularity.

On March 18, 2011, the exchange rate of Bitcoin to the US dollar hit a new low in seven weeks, dropping to $0.7.

On April 16, 2011, Time magazine published an article about Bitcoin, and Forbes' "Cryptocurrency" report was released.

On May 29, 2011, Rickard Falkvinge, the founder of the Swedish Pirate Party, announced that he had converted all his assets into BTC and borrowed a lot of money to accumulate BTC.

On June 8, 2011, after the exchange rate surpassed $10 on June 2, the transaction price on Mt.Gox peaked at $31.91, but four days later, the price plummeted to $10 and experienced a six-month decline.

On June 19, 2011, hackers stole user certificates from infected computers on Mt.Gox, leaking data from 60,000 users. The hackers then generated a large number of $0.01 orders to buy in, while the market price was $17.51, affecting accounts worth $8,750,000. The short-term abnormal price was declared invalid by Mt.Gox, which was forced to suspend trading for seven days to fix the bug. In the following months, multiple platforms were hacked, and the Bitcoinica platform ceased operations after being attacked twice.

On August 20, 2011, the first Bitcoin conference and expo was held in New York, and Bitcoin's attention reached a new high in Google Trends, with the price at $11.

On November 14, 2011, Bitcoin's price hit a six-month low of $1.99.

2012 Market

On March 1, 2012, the server's super administrator password was leaked, resulting in the theft of 46,703 bitcoins worth $228,845, a nightmare that haunts the Bitcoin world.

On July 22, 2012, users posted a total of 1 million posts on the official Bitcoin forum, with the price at $8.77.

On September 15, 2012, the London Bitcoin conference was held, with the price at $11.8.

On September 27, 2012, the Bitcoin Fund was established, with the price at $12.46.

On November 25, 2012, the first Bitcoin conference in Europe was held in Prague, with the price at $12.6.

On November 28, 2012, the block supply was halved for the first time, decreasing from 50 to 25 every 10 minutes, while the total issuance of Bitcoin reached half of the total supply of 21 million, with the price at $12.4.

On December 6, 2012, the first Bitcoin exchange operating under EU law—the French Bitcoin Central Exchange—was born, marking the world's first officially recognized Bitcoin exchange, with the price at $13.69.

2013 Market

In February 2013, the social news website Reddit, with an average daily PV of 40 million, announced that it would accept Bitcoin payments for its paid services.

On February 8, 2013, Coinbase claimed a monthly trading volume of 1 million bitcoins, with the price at $22.7.

On February 19, 2013, Bitcoin client version 0.8.0 was released, with the price at $28.66.

On February 28, 2013, BTC price surpassed the historical high of $31.91 from June 8, 2011, after 601 days.

On March 16, 2013, the government of Cyprus, an EU member state with a population of only 1.1 million, froze citizens' bank transfer transactions and imposed taxes on bank deposit accounts, closing banks and stock markets on the 18th, causing Bitcoin to rise to $47.45.

On March 18, 2013, the US Treasury's Financial Crimes Enforcement Network (FinCEN) released regulations for virtual currency management, clarifying the concept of virtual currency for the first time, with the price at $47.8.

On March 28, 2013, Bitcoin's total market value exceeded $1 billion, with an average price of $92.

On April 1, 2013, Bitcoin prices exceeded $100 across all trading markets.

On April 10, 2013, BTC reached an all-time high of $266.

On April 16, 2013, trading volume on the Mt.Gox platform reached a record high of 610,000, with a turnover rate of 5.8%. The price at that time was $71.76.

On November 3, 2013, the CEO of eBay praised Bitcoin in public and mentioned the possibility of considering Bitcoin as a payment method, with Bitcoin prices reaching $218.95.

On November 6, 2013, after a seven-month hiatus, Bitcoin returned to the $266 high, receiving extensive media coverage globally.

On November 15, 2013, Coinbase's app was listed on Apple's App Store, allowing US users to purchase Bitcoin on their iPhones.

On November 27, 2013, Bitcoin reached $1,078, surpassing $1,000 for the first time against the US dollar.

On November 29, 2013, Bitcoin touched a high of $1,240, equivalent to the price of one ounce of gold at the time, attracting global attention.

On December 5, 2013, five central government ministries jointly issued a notice on preventing Bitcoin risks, causing Bitcoin to drop from a high of $1,230 to as low as $870.

On December 7, 2013, Baidu announced it would stop using Bitcoin as a payment method, causing Bitcoin to drop to $697.

On December 8, 2013, a Lamborghini dealer announced support for Bitcoin payments for purchasing Tesla cars, causing Bitcoin to rebound to $972.

On December 16, 2013, the central bank interviewed third-party payment platforms, leading Chinese Bitcoin trading websites to cease all recharge channels, resulting in a significant drop in Bitcoin prices, with a single-day decline exceeding $200, and hitting a low of $454 on the 18th.

On December 20, 2013, some Chinese trading websites resumed recharges, leading to a rebound in Bitcoin prices.

On January 3, 2014, it was the fifth anniversary of the birth of Bitcoin's first block, and Bitcoin prices rebounded to $888.8.

POW Class

Proof of Work (POW) is, as the name suggests, proof of the amount of work done. Generally, it can only be proven from the result, as monitoring the work process is usually cumbersome and inefficient. The mechanism of proof of work may seem mysterious, but its applications in society are very widespread. For example, graduation certificates and degree certificates are forms of proof of work; possessing a certificate indicates that you have invested time in learning and work in the past. Most things in life are judged based on results.

The principle of proof of work is simple. One party (usually called the prover) presents a computational result that is known to be difficult to compute but easy to verify. By verifying this result, anyone can confirm that the prover has performed a certain amount of computational work. The first POW program was developed in 1996 by Adam Back in the "Hashcash" program, which uses SHA-256 proof of work measures to combat spam—requiring all received emails to use a strong POW attachment. This system makes it economically unfeasible for spammers to send large volumes of emails while allowing individuals to send messages to each other when needed. To this day, a similar program is Bitmessage, and this algorithm has taken on new significance as the core of BTC security in the form of "mining."

For cryptographers, SHA256 is known as a one-way function, meaning its result is difficult to compute but easy to verify. To reverse-engineer this result, one would need to try every input until a random number matches. The traditional output form of SHA256 is a series of hexadecimal 64-digit numbers and letter combinations taken from 0123456789abcdef. For example:

SHA256("hello")=2cf24dba...

SHA256("Hello")=185f8db3...

SHA256("Hello")=2d8bd7d9...

The output of SHA256 is designed to have high randomness; even a slight change in the input data will produce a vastly different output, which is why SHA256 is called a one-way function.

Finding an input value for SHA256 that starts with "0" on average requires about 16 attempts, while finding one that starts with "00" requires about 256 attempts, and so on. The way Hashcash and Bitcoin mining work is that users (i.e., email senders or miners) are required to find a "random number" such that SHA256 (information + random number) starts with a string of zeros, and then send the found random number along with the information as proof of work. For example, the block numbered 254291 is

000000000000003cf55c8d254fc97d2850547e5b787a936bc729497d76443a89

On average, performing calculations within a block requires about 72.957 trillion attempts to find a random number that returns such a result starting with many zeros (technically speaking, 282.394 trillion attempts, because the POW requirement is more complex than just "starting with many zeros," but generally speaking, this is the case). The difficulty is artificially set to prevent attackers from dominating the Bitcoin network, introducing another blockchain, and overturning previous transactions. Anyone wishing to attack the Bitcoin network must complete at least 282.394 trillion SHA256 calculations.

(1) SHA 256 Class

Since Bitcoin uses the SHA256 mining algorithm, the current BTC network hash rate has reached 7PHash/s, making it easy for other coins using the SHA256 algorithm to suffer a 51% attack. Therefore, there are currently few competitive coins using the SHA256 algorithm. Most competitive coins using the SHA256 algorithm adopt "merge mining," enhancing their network security by utilizing the hash power of the Bitcoin network. Competitive coins using the SHA256 algorithm include NMC, TRC, IXC, DVC, FRC, etc., among which NMC, IXC, and DVC use merge mining, while TRC and FRC are completely independent of the Bitcoin blockchain.

1. 1. Terracoin

http://terracoin.org/

Terracoin (TRC) was released on October 26, 2012, with a total coin supply of 42 million. A block is generated every 2 minutes, slightly faster than LTC, with 20 TRC per block, halving every four years. TRC uses the SHA256 algorithm and operates on a new blockchain independent of Bitcoin. Due to TRC not adopting merge mining, its hash power is relatively low, and it was once subjected to a 51% attack, forcing it to frequently adjust its difficulty adjustment strategy.

(2) Scrypt Class

LTC uses the scrypt algorithm to effectively enhance blockchain security, becoming the second-largest cryptocurrency after Bitcoin. Various competitive coins have adopted the scrypt algorithm and improved versions of scrypt to increase blockchain security. Currently, there are over 100 competitive coins using the scrypt algorithm, among which notable ones include LTC, FTC, WDC, MEC, YAC, etc.

1. 1. Litecoin

http://litecoin.org/zh_Hans

Litecoin (LTC) was released on October 7, 2011, and is Bitcoin's closest follower. In its own words: if Bitcoin is gold in the currency world, we are silver. LTC is currently the highest market cap competitive coin.

Compared to Bitcoin, Litecoin has the following characteristics:

Faster issuance speed. Litecoin's issuance speed is an average of one block every 2.5 minutes, four times that of Bitcoin, allowing for six confirmations in 15 minutes, facilitating merchant transactions.

The expected total issuance is four times that of Bitcoin, meaning the final number of Litecoins will be 84 million.

Bitcoin is a grand social experiment, and many competitive coins are trying to solve various problems, exploring different development directions for cryptocurrencies. Regardless of success or failure, some competitive coins have indeed provided us with valuable insights and contributed to the advancement of digital currency.

Payment Systems and Decentralized Exchanges#

Dilemmas of Centralized Exchanges

Whenever there is a significant market fluctuation, exchanges are often the first to encounter problems. Whether it was the early Mt.Gox or the current Bitcoin China, this seems to have become a routine in the Bitcoin industry. How to solve this problem has been a persistent challenge for the entire Bitcoin industry. Currently, traditional centralized digital currency exchanges appear to have become a bottleneck and obstacle in the development of digital currencies.

(1) Government Regulation Issues

Centralized exchanges are easily subject to various problems due to government and bank regulations. Whether it was Mt.Gox in Japan or Bitstamp in Europe, both have repeatedly adjusted policies due to government or bank regulations. Mt.Gox, in particular, had $5 million in its bank account seized by the US government due to financial regulations, which led to withdrawal difficulties and caused significant market turbulence.

Government regulation issues hang over the market like the sword of Damocles, keeping the market on high alert. Any news related to government market regulation tends to cause fluctuations. Since both the US and Chinese governments have maintained an unclear attitude towards digital currencies, market participants have many speculations about the potential attitudes and measures the government may take. This is certainly not a favorable environment for the development of digital currencies.

(2) Website Technical Architecture Issues

Due to the rapid development of digital currencies, many technical personnel entering the digital currency trading market often lack sufficient preparation. We can observe that both domestic and foreign exchanges have weak technical architectures, whether in software or hardware. When instantaneous trading suddenly surges, exchanges often experience significant delays or even complete outages. Moreover, these issues typically occur during periods of market volatility, which can have a tremendous impact on the market. Although many new exchanges have emerged over the past year, reducing these impacts significantly, if major exchanges experience delays or outages, it will still have a significant effect on market conditions.

Additionally, due to the astonishing speed of digital currency development, a massive influx of new users has occurred. Despite many websites continuously updating, they often still cannot meet user and market demands, leading to significant complaints and dissatisfaction with trading websites. Market surveys have shown that the vast majority of market participants have some level of dissatisfaction with trading websites. Currently, it seems that no single website has emerged to demonstrate outstanding capabilities in providing excellent services to the market.

(3) Website Integrity Issues

Due to strict regulations on websites abroad, the development of trading markets in Europe and America has been relatively slow compared to domestic markets. However, the relaxed regulatory environment in China has often resulted in a mixed bag of domestic exchanges, the most notorious being the GBL digital currency futures exchange, which absconded with funds. Although this case did not significantly impact the market and many signs were evident before the escape, with relevant suspects already apprehended, it still highlights the significant integrity issues within the domestic trading market.

Building a barely usable trading website often does not require high costs, but a large amount of funds can remain on the website. If the website manager has malicious intentions, it could cause significant losses to users. Furthermore, as mentioned in the second point, if the website's technical architecture is not well-developed, once trading difficulties or withdrawal issues arise, or even a complete outage occurs, a flood of negative rumors about the website can emerge in the market. Although most of these rumors are proven to be unfounded, they still create significant psychological pressure for many newcomers. Thus, in a market environment where participants cannot feel secure in trading, users always harbor anxiety, which is detrimental to the healthy development of the market.

If the issues surrounding centralized exchanges cannot be effectively resolved, they will undoubtedly add more uncertainty to the already convoluted path of digital currency development. Many people have turned their attention to Bitcoin itself, as Bitcoin solves issues like double spending and proof of work through decentralization. So, can decentralization also resolve the series of problems caused by centralized exchanges? Many have begun exploring decentralized solutions, leading to many excellent proposals. These proposals are primarily centered around decentralization, and we will later showcase their advantages and disadvantages.

Digital Assets#

(1) Digital Currency

Digital currency is a form of currency that exists only in electronic form within the internet environment. Digital currency can be issued by individuals, organizations, companies, or merely a software system. To date, digital currencies are not issued by national central banks and have not been directly recognized by any central bank in the world. Digital currencies do not necessarily need a benchmark currency. Digital currencies can be broadly divided into centrally issued currencies and decentralized currencies. The difference between the two is that centrally issued digital currencies are controlled by a specific person, institution, or company, and the issuance volume and method can be changed at any time based on demand. Examples of centrally issued digital currencies include Q coins, e-gold, and XRP in the Ripple system.

Decentralized digital currencies (Cryptocurrencies, as they are often created, distributed, and maintained using cryptographic and verification technologies to ensure no double spending occurs) are defined by their initial issuers (which can be individuals, organizations, or companies) regarding factors such as issuance volume, method, and speed. Once issued, important parameters like issuance volume and method can no longer be changed. Representative decentralized digital currencies include Bitcoin (BTC), Litecoin (LTC), and PPcoin.

(2) Digital Assets

Digital assets currently do not have a strict definition; generally, they refer to any valuable asset that exists in electronic data form. By this definition, all currencies, stocks, and bonds stored on computers should be considered digital assets. However, this definition can lead to ambiguity. For example, if a company's stock exists in both paper and electronic forms, if the electronic form is completely deleted, it does not affect the stock's value, which remains objectively existent. This means that the electronic form of the stock on a computer is merely one form; its objective value is unaffected by its existence, and whether its electronic form has true and independent value seems difficult to confirm.

Thus, I define digital assets as any assets owned by enterprises, institutions, or individuals that exist solely in electronic data form and possess independent commercial or exchange value. Digital assets can be divided into two main categories based on their form: standardized digital assets and non-standardized digital assets. Standardized digital assets have clear prices and forms, and after being divided based on their units, each portion should have the same value, allowing for clear measurement, such as Bitcoin, stocks of Cat Baking Company, Q coins, or Shanda's point cards. Non-standardized digital assets, when divided based on their units, do not have uniform value, and no unified algorithm can accurately indicate the independent value of each portion. Examples include domain names, QQ numbers, and email addresses. Although each of these assets is independent, the actual value of two identical types of assets may vary greatly.

Before the explosive development of digital currencies, even standardized digital assets were relatively singular and often had a close and fixed exchange relationship with fiat currencies. However, with the rapid development of digital currencies over the past two years, a variety of digital asset products, such as stocks, futures, and funds, have emerged in the form of digital currencies. We believe that in the near future, we will see more digital asset products that either already exist in traditional financial markets or have never appeared before.

In the past decade, digital assets have primarily been non-standardized. Some have invested in digital assets such as domain names and QQ numbers. Although these total amounts are quite small, they have become an indispensable part of life, with everyone owning some digital assets, such as QQ numbers, email addresses, diamonds in Taobao stores, blogs, and Weibo. We have also seen cases of inheritance and transfer related to these digital assets. However, compared to standardized digital assets, this remains a small portion. With the development of digital assets, we believe that their proportion in people's lives and the financial sector will continue to grow. In this context, digital asset management will become increasingly important.

(3) Digital Asset Management

Digital Asset Management (DAM) refers to the entrusted management, utilization, and disposal of digital assets, including digital currencies and equity, debt, options, and other property rights supported by digital currencies, to achieve goals such as preservation and wealth creation. The market for digital asset management is primarily reflected in the fund industry, entrusted wealth management business, and trust industry.

As the various business forms of digital assets gradually enrich, the activities and value-adding links created by various institutions related to digital assets are becoming clearer, and the liquidity and scale of digital assets will gradually increase. By allocating digital assets based on different value activities and their ecological positions within the digital asset industry, the unique core competitiveness of digital asset management will be reflected.

Digital asset management exhibits clear characteristics of cross-domain, cross-industry, and cross-technical levels compared to other asset management. It requires not only rich experience and concepts in financial product management but also a comprehensive understanding of the technical aspects of digital assets, enabling the flexible use of various digital asset-related technical tools. More importantly, since digital currencies are still in their early stages of development, it is essential to have an in-depth understanding of the entire digital currency industry and to grasp the trends and processes of digital currency development.

As a component of the financial system, the digital asset management industry, while differing in business links and forms from traditional industries, generally includes the main stages from product research and development innovation, channel layout, internal organizational management of enterprises, to customer service experience. Specifically, upstream businesses include asset allocation, investment business, investment advisory, and product design and innovation; midstream businesses include providing channel services and product maintenance; downstream businesses include developing financing clients, maintaining and servicing clients, and marketing and value-adding of brands.

Another characteristic of digital assets is their ability to be quickly transferred and circulated globally. Traditional fiat assets, securities, or precious metals face numerous restrictions on rapid transfer and exchange across the globe. However, in the face of rapidly changing economic and political situations in various countries, the high liquidity and convertibility of assets are aspects that many high-end individuals value greatly. When assessing asset liquidity, four indicators must be considered: spatial indicators, time indicators, cost indicators, and security indicators.

With the rapid development of digital technology, our living space has expanded from the physical realm to the virtual realm. In this context, the concept of asset transfer and exchange has evolved from a singular notion to encompass two levels: the ability to quickly transfer one's assets globally across countries and regions, and whether assets can be rapidly exchanged in both online virtual spaces and offline physical spaces. In this situation, the unique advantages of digital assets are once again highlighted. For digital assets that inherently exist on the internet, there are no geographical limitations; they can be rapidly transmitted wherever the internet and mobile internet reach. Additionally, the emergence of digital trading venues and decentralized digital asset trading platforms worldwide allows those in need to quickly exchange digital assets in both virtual and real-world contexts. Traditional assets, on the other hand, face many restrictions and various regulatory controls.

In terms of time, cost, and security, digital assets possess unparalleled advantages. This aspect does not need further elaboration; anyone with a basic understanding of digital assets should be very clear about this.