The ten-year history of mining industry (Part 2) - Mining machines will eventually come to an end, but mining will last forever

Let’s go back to the story in 2013 in the previous article.

There has always been a saying that when doing business, "those who dig for gold are not as good as those who sell water". In this year's blockchain industry, we have seen countless examples of this: many powerful teams did not become project parties, but chose to become institutions that serve project parties, such as media, exchanges, and market value management. But strictly speaking, this is not a new idea. As early as around 2013, this kind of thinking was already popular in the mining industry. The original intention of its derivative is that many people found that although being a miner was quite popular in the currency circle at that time, it was not as "profitable" as the legend said. On the contrary, the risk was actually very high - in terms of income, the amount and price of coins were either uncertain or fluctuated greatly. In terms of cost, it is difficult for digital currency mines to negotiate direct power supply contracts with power plants as large users, which leads to high electricity prices. Under such circumstances, these people focus on the upstream supplier link - that is, mining machine manufacturing and supply.

Driven by this kind of thinking, after the bull market in June 2013, social capital that intended to enter the mining industry no longer took mining as the sole goal, but split into two groups. Some people rushed into the downstream mining field to become miners, while the other part entered the upstream mining machine manufacturing industry as suppliers. For a time, there were hundreds of mining machines competing in the mining machine field. In addition to the more familiar Friedcat and Canaan Creative, many other brands appeared in the market, such as Dove, TMR, Biter, Land, Little Bee, Garden, Smart, Ant, ASICME, Rain Edition, MQH, Dragon Mine, Shenyu, 42BTC, Shaoming, FBT, Jingchen, Krypton, Zeus, West, etc.

Figure: Many mining machine manufacturers and OEMs that were once active now can’t even find a complete logo

It can be said that the number of mining machine manufacturers around 2013 was comparable to that of current blockchain media and exchanges. However, the competition between them was not as bad as that of current market players. On the one hand, most people who were capable of mining machines were geeks and technology geeks, who were relatively straightforward and simple and not good at undermining each other. On the other hand, the most time-consuming and energy-consuming aspect of mining machine companies at that time was not the friendly competitors within the industry, but the ups and downs of the external environment.

Specifically, in the second half of 2013, when many entrepreneurs in the mining machine field had not entered the industry for long, the price of Bitcoin once again rose wildly, from about $40 in the middle of the year to $1,200 in November 2013! Instantly, the society was in turmoil, and many people stared at the increase in Bitcoin and fell into anxiety. Among them were not only traditional economists and financial scholars who could not understand this new thing, but also the various mining machine manufacturers mentioned above. The specific reasons have been mentioned many times by our superiors: the rise in currency prices is inevitably accompanied by an increase in computing power, which will increase the difficulty and reduce the miners' yield. For mining machine manufacturers, this means that in order to meet customer needs, they have to upgrade their product performance again, that is, compress the line width of mining machine chips, increase computing power, and reduce power, thereby increasing computing power density, so as to outperform the computing power scale of the entire network as much as possible.

It is no exaggeration to say that in the manufacturing industry, mining machine manufacturers may be the most exhausted type. If we carefully observe the updates and iterations of various technological products, it is not difficult to find that for traditional electronic products such as mobile phones and computers, their update and replacement cycles are often in years, such as Apple released the iPhone 4 in 2010 and the iPhone 4s in 2011; while in the field of mining machines, this cycle is shortened to months, such as Roast Cat launched the spot 13G blade mining machine in July 2013 and the spot 10G Mini mining machine in August 2013. There is really no way to deal with this situation, because in the early days when there was only Bitcoin mining, due to the rapid growth of computing power and the short product update and iteration cycle, everyone was forced to rush forward by the industry. Often, after receiving the advance payment, the company had to immediately invest in the development of new models, otherwise it would be squeezed out of the market in minutes.

Table: Changes in Bitcoin network computing power, which often increases exponentially within a year (Unit: Hash)

Data source: Jingzhun Research Institute

From this point of view, in the previous article, we compared mining machine manufacturers to arms dealers, which is quite appropriate. At least their ability to burn money is comparable. Fortunately, the booming digital currency market and the situation of supply exceeding demand at that time gave mining machine manufacturers the funds and motivation to continue research and development, so that everyone ignored a problem - that is, what if the cash flow is cut off one day?

This seemingly impossible assumption became a reality in December 2013. In that month, the People's Bank of China and five other ministries and commissions issued the "Notice on Preventing Bitcoin Risks". The price of Bitcoin was instantly cut by 50%, plummeting to $600. The mining industry's enthusiasm instantly cooled down, the market capacity shrank sharply, and the mining machine companies that lost their capital infusion also fell into trouble. On the one hand, the old inventory machines could not be sold, and on the other hand, there was no money to develop more powerful models. In this way, they actually faced only one way out - closing down. Therefore, in the dark period when Bitcoin continued to fall to $200, a large number of mining machine companies mentioned above disappeared from people's sight one after another, and even the leading boss, Fried Cat, ran away and lost contact, completely evaporating from the world, and the industry was in mourning for a while. This situation must be familiar to readers who are in the current bear market, so I will not go into details here.

Strictly speaking, the plunge of Bitcoin is certainly influenced by policy, but as a financial asset that follows the economic cycle, its bear market is inevitable. Of course, from another perspective, it can also be understood that the arrival of the bull market is inevitable. The budding of the bull market was born in the second half of 2015. After hitting a low of $210, Bitcoin began to rise slowly. The dormant miners also began to become active and prepared to regroup and start the battle again. However, what surprised many old miners who returned to the battle was that the market at this time no longer had the grand scene of hundreds of machines competing in 2013. Looking around, there were only two relatively powerful suppliers left: one was Canaan Creative, and the other was Bitmain, which was established at the end of 2013.

Figure: The last two giants surviving in the mining machine field after the bear market in 2014

As a veteran miner with rich customer resources, it is not surprising that Canaan Creative survived the industry's cold winter. In fact, many people are quite interested in the following point: Among the two giants of mining machines, how did the other rising star, Bitmain, survive the bear market and become the leader in the industry? In this regard, there is no reliable information available on the Internet. Public information shows that by continuously iterating its Antminer four times in 2014, Bitmain went against the trend in the bear market, eliminated all its competitors, and became the leader in the mining machine industry. But the problem is: everyone knows that it is a good thing to buy at the bottom of the bear market, but the key is where did Bitmain get the funds for research and development? You must know that most venture capitals at that time were running around Bitcoin. One possible factor is that Bitmain used its mining pools and mines to provide blood transfusions for its upstream main business. According to the recollections of Bao Erye, a popular person in the currency circle, Bitmain once bought mines from him at a low price. Although the specific details are unknown, in short, through a series of arrangements, Bitmain and Canaan Creative eventually became the few survivors of the 2014 bear market.

At this point, after the downstream mining link has formed a pattern of several major mining pools running side by side, the upstream mining machine link has also changed from a group of heroes to a two-strong competition, and it can even be said that there is only one leader. After all, compared with Canaan Creative, Bitmain's leading position is actually indisputable.

The most feared result comes when you rule the world

Strictly speaking, Bitmain's dominance is not the company's subjective intention. To put it in a broader sense, it is an inevitable development of the mining machine industry. After all, only a few market players can survive the bear market, and the bear market is an inevitable component of the market. But no matter what, for the mining industry, it has entered a situation that people don't want to see - that is, both upstream mining machines and downstream mining pools have begun to become very centralized.

Let’s first look at the mining machine link.

As Bitmain controlled the supply rights of most POW digital currency mining machines represented by Bitcoin around 2015, the company has already possessed a "dominant" position in the mining industry. It is no exaggeration to say that if Bitmain sneezes, the entire mining industry will be affected. Unfortunately, at this cusp of the storm that the entire industry is watching, Bitmain has repeatedly given people a handle and delivered multiple shells to its opponents in the public opinion field - first, it took advantage of the loopholes in the Bitcoin algorithm to develop an ASIC Boost program, making its own mining pool 20% faster than its competitors. Later, the "Ant Blood" scandal occurred, that is, it was revealed that there was a security vulnerability in its mining chip, allowing Bitmain to remotely shut down the mining machine by just executing a few lines of code. All kinds of "uneasy" behaviors have caused discussions in the industry for a while.

If the above two events can be regarded as small-probability events with a high degree of chance, and do not have much lethality in today's world where netizens' memories are only seven seconds, then for miners, they have continuously and personally experienced the pain of the seller's market brought about by the centralized monopoly of mining machine manufacturers in the daily process of purchasing mining machines.

As mentioned above, the iteration cycle of the mining machine industry is very fast. For miners, they often need cash flow to quickly develop new products to chase the rising computing power of the entire network. In order to achieve this goal, selling futures in advance to obtain cash has become a common practice in the industry. However, the problem is that since Bitmain almost monopolizes the mining machines of major currencies, it not only sells futures when selling on the official website, but also requires full payment. The delivery date is generally about two months later. As for the risk of fluctuations in the price of the currency and the computing power in the middle, I'm sorry, the buyer bears it by himself. If the machine obtained after the delivery cannot outperform the computing power of the entire network, then you can only touch your nose and admit your bad luck. After all, Bitmain covers the widest range of currencies, and some currencies are even the only one. For such a supplier, you users can buy it or not, and you are not much more, and you are not much less.

After looking at the mining machine part, let’s take a look at the mining pool.

As we mentioned in the previous article, after ASIC chips landed in the mining field, they relied on their overwhelming computing power advantage to launch a continuous charge towards all POW mechanism digital assets. The issuance rights of major digital currencies have also been continuously concentrated in the hands of small groups. Projects that do not have built-in ASIC resistance (such as Litecoin and Dash) are needless to say. They were attacked by ASIC shortly after their debut. And those currencies whose mining algorithms have built-in ASIC resistance have also been broken one after another after several years of hard battles. Take Monero, Ethereum, and ZCASH as examples: In March 2018, Bitmain stated that it would launch Monero ASIC mining machines; in April, it confirmed that it would launch Ethereum ASIC; in May, it announced that it would launch ZCASH ASIC mining machines. After these three shots were fired, miners in major cryptocurrency communities were in a panic for a while. Among them, Monero even launched the shadow clone technique on the spot and split it into five, and did not hesitate to use forks to fight against the established ASIC mining machines.

The strong centralization of mining machines and mining pools not only brings a lot of inconvenience to the miners involved, but also makes people outside the blockchain industry use this as an excuse to criticize the decentralization of blockchain as a false proposition. In fact, when we look at the real economic society, it is not difficult to find that the combination of "two giants of mining machines + many giants of mining pools" is very similar to a super-centralized industry that people are familiar with in the real world - that is China's power sector, and its industry structure is actually "many giants of power plants + two giants of power grids". Behind this terrifying analogy, the information revealed is that the current degree of centralization of traditional mining has reached the level of energy central enterprises, and it has nothing to do with the word "decentralization". As for the "Ant Blood" vulnerability mentioned above, even if it is an unintentional act of Bitmain, its appearance in the future is actually just a matter of time - as long as the mining machine manufacturer wants to shut down a mining pool, it will not let you mine more for a minute; just like the power grid company wants to limit the power supply of a power plant or user, it will cut off the power supply for you in minutes.

Figure: The degree of centralization of the mining industry is comparable to that of China's power industry, which is dominated by central enterprises.

Undoubtedly, this is a highly controversial topic. Although some people believe that the increase in the degree of centralization of mining is the result of social competition and division of labor, and use the economics of the real world as evidence, for believers of "decentralization", they come to the digital world because they are dissatisfied with the centralized structure of your real world. If you use reality to make a point, how can they buy it: since the centralization of traditional mining is the inevitable result of the development of your POW mechanism, then why don't I just not play with you? Therefore, even though the supporters of "centralized mining" have many reasonable reasons to support their arguments, most investors still vote with their feet - that is, they choose to participate in digital currencies with consensus mechanisms other than POW.

Giants exit: Mining machines are over, but mining will last forever

At the end of 2016, when the global financial market was in turmoil and stocks, bonds, currencies and futures were wiped out, a large amount of funds suddenly poured into the digital currency market of Avalon, an isolated island. Affected by the Saigon effect, a new round of bull market suddenly arrived without any warning. Compared with the previous rounds of bull market, this bull market has an unusually obvious feature: if the three bull markets in 2010, 2011 and 2013 gave birth to GPU, FPGA and ASIC technologies into the mining field, then the bull market in 2017 did not bring about a revolutionary upgrade of equipment, but instead some improved designs, such as further compression of chip line width.

Why does this happen? We have given the answer above. In addition to the development limitations of chip technology itself, there is another important reason. With the introduction of the concept of blockchain and related terms, more and more practitioners have jumped out of the circle of digital currency and realized one thing: the achievement of distributed ledger consensus and the distribution of tokens do not necessarily rely on the "guessing contest"-style POW mechanism. Moreover, under the pressure of energy environment and performance constraints, a single POW mechanism is no longer the key development direction in the future. Even if it is adopted, it may be in the form of a hybrid consensus mechanism.

Table: Impact of the four major bull markets in history on the mining industry

Under such circumstances, although thousands of new projects have emerged in the digital asset market in the past year or so, there are only a few projects that simply use the POW mechanism (especially well-known mainstream projects). At least their growth rate has not kept up with the growth rate of the digital currency industry itself. Moreover, those projects that seem to have a brighter future have also clearly stated that they will abandon the POW consensus mechanism. For example, the Ethereum project, which is generally regarded as blockchain 2.0 in the industry, has clearly stated that it will switch to the POS mechanism, while the "Blockchain 3.0" EOS simply did not use POW as a transition, but directly used the DPOS mechanism.

As a result, the traditional mining industry, which consists of mining machine manufacturers and mining pool operators, faces a very scary future prospect: the road ahead of them, like the POW mechanism tokens whose market value accounts for an increasingly smaller proportion, is actually getting narrower and narrower - whether it is to transform the existing market, that is, to further optimize the computing power density of existing chips; or to capture the incremental market, that is, to continue to break the resistance of other POW tokens, in the foreseeable future, they will encounter a bottleneck of increasingly smaller marginal effects. After all, the line width cannot be infinitely compressed, and there are only so many types of POW mechanism tokens on the market now, and there will be no more after the research and development. In other words, the traditional mining industry that uses high-computing equipment represented by dedicated chips to mine is actually heading towards dusk.

Mining machine manufacturers are actually well aware of this, so everyone can see that although they don’t say it, the leading mining machine players including Bitmain and Canaan Creative have actually begun to carry out corresponding follow-up work in the following two aspects since the beginning of this year:

First of all, in terms of the capital market, they are eagerly seeking IPOs. For the blockchain industry, which is naturally resistant to traditional finance, this approach is a bit like "reversing the times." However, perhaps only the helmsmen of these companies themselves understand that if there is no huge bull market in the next two to three years, then the first quarter of 2018 will become their peak moment, just like the 6124 points of A shares. If not take advantage of this time to estimate a good market value and cash in the equity in hand, when will it be?

Secondly, in terms of business operations, the two mining machine giants have also begun to seek new business breakthroughs, such as transforming into artificial intelligence chip suppliers in the fields of smart home, autonomous driving, and image recognition. Compared with the simple trial of AI in the last round of bear market, the transformation of mining machine manufacturers this time can be said to be more thorough and resolute, and they even spend a lot of real money on businesses that cannot bring cash flow for the time being, such as producing some products that are not very competitive for the time being, as a future technology accumulation. Regarding this, some interpretations believe that it is the response of chip manufacturers to China's national strategy under the background of the Sino-US trade war, but the author personally believes that, apart from the factor of self-whitewashing due to the large size of the enterprise, this is actually a self-help behavior of mining companies in the current blockchain industry environment, which has little to do with the international situation. After all, if you calculate carefully, the time for mining machine manufacturers to transform AI chips is actually much earlier than the trade war.

Table: Specific ways in which mining machine manufacturers carry out follow-up work

Figure: Bitmain released its first AI chip, “Suanfeng”. Although its performance is not very powerful, it is still the first step in the field of AI.

However, although the traditional mining industry has shown signs of gradual decline, when we observe the entire blockchain industry, we will find a very interesting phenomenon, that is, the word "mining" is still active in the industry, and the level of activity is still increasing. In fact, in the field of blockchain, the word "mining" has long exceeded its original meaning and has become synonymous with "obtaining digital currency". For example, the popular Gongxinbao data mining and Bihu writing mining some time ago, although they are all claimed to be "mining", if you look closely, you will find that the token output mechanism of these projects has nothing to do with the narrow sense of POW mining. In addition to basic infrastructure such as mobile phones and computers, you do not need any special equipment. For example, the currently popular IPFS mining machine is assembled using general electronic equipment such as CPU, memory, and hard disk. You may not even need equipment. For example, the "transaction mining" that was highly sought after some time ago is to distribute tokens according to the proportion of the user's transaction amount on the exchange client.

As a result, the blockchain industry may see such a scene in the future: perhaps one day, facilities and organizations including mining machines, mining farms, and mining pools will no longer exist in this industry, but the word "mining" is likely to be passed down and become a very important but unclear term in the blockchain industry, because the future "mining" is not only completely different from the past Bitcoin mining, but even we may not be able to imagine it now. After all, don't forget: when we lament that "an old era has ended", we forget the other meaning behind this sentence, that is, a brand new era is coming.