Yang Xin, Product Director of Bit Continental: The mining technology threshold is much higher than the funding threshold

On the afternoon of March 29th, Bitland announced that the new S17 series of ant miners equipped with the second generation 7nm chip will be officially released on April 9th ​​and will be available from stock when released.

Compared with the previous two months after the release of the mining machine, Bitcoin has made a lot of efforts to meet the needs of Fengshui mining.

Just five hours after the announcement of the post-birth official announcement, Jia Nan Zhizhi also announced that Avalon's new bitcoin mining machine A10 series will be unveiled.

As in other chip areas, the competition for mining miners is also a race against time. A little bit worse, the market is likely to be occupied by the opponent's excellent new products.

• Do you understand the Plasma framework? It’s all here.
• The copy of the private key is stolen, the address transfer is anti-lost, and the data security of the blockchain C is guaranteed.
• Getting started with blockchain: Why should Ethereum set a block for the block?

Looking back at the world's more than ten mining machine manufacturers, there are still a lot of people standing still. "All competition failures are not as good as people", and Yang Xin, the product owner of Bitland, understands this.

In the face of the mining machine production of this oil field, there are constantly new players entering the field to try. However, Yang Xin believes that talents in the field of computing chips are also scarce in the world, and the talents for developing chips for digital currency are even rarer.

"Talent is the most expensive, and the technical threshold of this line is much higher than the funding threshold." Yang Xin said. In addition to a mature product chain, Bitcoin also has high-end R&D talents. This may be the moat on its way to the Iron Throne.

On the occasion of the release of the new S17, Odaily Planet Daily made an exclusive interview with Yang Xin. In the interview, he showed us the process and potential value of the current advanced 7nm chip development and the technical concept of Bitland. Of course, I also answered various questions about S17.

The following is an interview record, enjoy:

S17 makes money faster

Odaily Planet Daily: How long have you been to Bitland, and what are you responsible for now?

Yang Xin: It’s almost three years. I was the first to be responsible for embedded software work in Bitland. Later, he gradually participated in the development of the software of the mining machine and the entire product, and finally was responsible for all product delivery. From the earliest chip definitions to chip design, verification, manufacturing, certification, and even customer support, I am responsible for the entire process. (Note: Embedded software is the operating system and software embedded in the hardware)

Odaily Planet Daily: What are the characteristics and advantages of S17?

Yang Xin: First, the energy efficiency ratio is low. Now, for Bitcoin, the proportion of electricity bills is already high. Therefore, this sharp drop in energy efficiency ratio means a significant drop in costs, and the cost reduction is a profit.

Second, the performance is strong. Regardless of the energy efficiency ratio or the unit volume of the calculation, the new machine has been greatly improved, which is very important for people who invest heavily in the mine. The calculation of unit volume means that some mining machines can do a lot of calculation, but the bulk of the same as the refrigerator, purely two piles for one, then its calculation capacity is not as good as S17.

Odaily Planet Daily: It is based on the S15 series to optimize, is it based on the S15 to iterate?

Yang Xin: As a second-generation 7nm chip, it is obviously a new iteration. There is still a generation difference (which can be understood as a qualitative leap), not a simple optimization.

Odaily Planet Daily: How long is the S17 series development cycle?

Yang Xin: This is hard to say. Because many of our R&D is rolling research and development, because the technology is a combination of more than a dozen technologies, some of them need to be invested very early and some need to take some time to develop later, some problems even need We must solve it within ten days.

Odaily Planet Daily: The most mainstream on the market today is the cost-effective S9. Will there be not many people buying S17?

Yang Xin: No. In 2016, or last year, buying S9 is definitely the most cost-effective; now it is definitely worth buying new ones. It is indeed faster to make money.

7nm only TSMC do

Odaily Planet Daily: Mining chips, is the higher the process, the stronger?

(Note: The process refers to the fineness of the integrated circuit. Currently, the "nm" is the unit. The higher the accuracy, the lower the "nm" value and the more advanced the production process.

Yang Xin: In simple terms, when the process is more advanced, it will bring huge advantages in performance, but it is also more expensive.

Odaily Planet Daily: Is the price high in research and development costs, or is it limited in capacity?

Yang Xin: There are several reasons: First, technical difficulty. The development of the process is actually getting slower and slower because it has physical limits. At 7nm, this quantum effect is already very obvious.

Second, the difficulty of breaking through the entire industrial chain has led to insufficient capacity. From the earliest 130nm chips to advanced technology, there are more than a dozen factories around the world. However, when it quickly evolved to 55nm and 40nm, the number of players involved decreased dramatically. why? Because the development costs of each generation are rising sharply. If the manufacturer can't make a profit from it, it will give up this technical competition.

By the time of the 7nm generation, there is basically only one in the world, that is, TSMC has realized this process. The old Intel and IBM have no mature technology in this respect. This is a very big reversal. Before that, Intel and IBM were still very powerful.

Odaily Planet Daily: What are the widely used products of 7nm chips in daily life?

Yang Xin: In fact, it is not widely used. You may only see the most advanced Apple phones or other high-end phones, and some of the core chips may be used. NIVIDIA graphics cards, or AMD's CPUs, may also be used. Because the production capacity of 7nm is only one of TSMC's, it is in short supply, and it has not yet reached the point of “big cabbage” everywhere.

Odaily Planet Daily: Does it hinder the large-scale use of such chips, are the problems mentioned above?

Yang Xin: Yes, 7nm The problem now is mainly due to the capacity of TSMC. Everyone wants to use it, but TSMC's production line has already run full, and there is no second home to do. This one, there is no branch, so I don’t want to use it.

We and TSMC are investing 5nm technology.

Odaily Planet Daily: A hardware product has different stages. From early experiments, maturity, to large-scale application, your 7nm development to the second generation, you judge which stage is it?

Yang Xin: The level of our products is suitable for large-scale promotion and sales. It is experienced in the process from trial production to small batch to mass production. We don't have experimental products for everyone, this is impossible.

Odaily Planet Daily: 7nm next-generation chips still have the power to develop?

Yang Xin: From a technical point of view, there is no end. Its next generation is called 5nm technology, and we have been working on R&D as early as TSMC. But a technology starts from research and development and can be put into use. It is an annual investment. Depending on the situation, each generation is different. Some key components have to be laid out ten years in advance. Now 5nm, and the next-generation updated technology, has been researched in TSMC or upstream and downstream manufacturers, and 7nm technology is being carried out at the same time. But when it can be used for mass production will be slower and slower. Therefore, in the near term, 7nm should still be the most preferred.

The real hegemony of the computing chip is just a few

Odaily Planet Daily: Which mining field is the bit of the continent's mining chip?

Yang Xin: There are three basic industries in IT technology: storage, communication and computing. Bit-continental chips belong to the field of high-performance computing.

Odaily Planet Daily: Where are the main applications of computing chips, and what are the more well-known vendors in this field? How many mining chip manufacturers are there?

Yang Xin: There are so many masters of computing. Intel, AMD and NVIDIA, many of China's supercomputers use their chips. Bitcoin's supercomputing should narrow the field a bit more, it is a special super-calculation for digital currency. Like the manufacturers of mining chips, there have been more than a dozen or so in the world.

Odaily Planet Daily: Can you give a general introduction to the chip development process?

Yang Xin: The process of chip development, simply said that after several stages, logic design, module design, layout design, after the completion, it will be handed over to TSMC. The logic design is based on the algorithm you want to achieve, and when it is implemented into the module, that is, the physical design, it has already been converted into a gate-level circuit and can be implemented in the space structure. (Note: gate-level, "gate" means logic gate, which is the lowest unit in the logic design. The circuit at this level needs to be combined with physical design.)

But it must be combined with the specific process of the foundry. The same design, produced in different fabs, is not the same as the final product. After the process is combined, the layout is formed, which we call a mask. The mask can be used for production, and it can also be called tape-out.

Odaily Planet Daily: After the chip was produced?

Yang Xin: After the TSMC production and processing, the chip went to the packaging and testing factory to seal the test chip, and finished the patch to the foundry (that is, the electronic components such as resistors, diodes, and chips were attached to the circuit board), and then assembled at our factory in Shenzhen. Mineralizer.

“Technical thresholds are often much higher than the funding threshold”

Odaily Planet Daily: You said that there are 20 manufacturers in the world before and after the mining machine. What are the main problems of the mining machine manufacturers who are doing the disappearance?

Yang Xin: The mining machine industry, or the blockchain industry, has been a global circle industry since its inception. To put it bluntly, all competition (failure) is not as good as people. This skill is not as good as people, not only your chip design, but also your mining machine design, but also your supply rhythm, even if you can smoothly send the mining machine to the customer and set them up, This is your "technique". It must be comprehensive in order to be in this market. Because the winner is all-in-one, in any industry, when it develops to a certain stage, there is only one head player with two or three companies that can survive, and then the rest will be abandoned by the market.

Odaily Planet Daily: I have seen the development of the memory industry in the past 20 years from the 40 to 50 homes, and now there are only a few years of mining chips. Do you feel that you have reached this stage?

Yang Xin: Almost.

Odaily Planet Daily: Even if there are not many manufacturers, we can still see some new miners joining, although their products have not been released, but there will be wind.

Yang Xin: Yes, there will always be some people who want to test, try once and for all. It’s always only after trying it.

Odaily Planet Daily: How high is the threshold for this industry?

Yang Xin: This threshold is crafted on the one hand. Behind the integrated circuit design technology, there is complete production, sales, delivery, etc., a whole product chain is the threshold. The second aspect is talent. Even in China, where the IC industry is strongly supported, such talents are difficult to recruit. The Bitcoin chip is a chip in the field of computing. This field has long been dominated by Intel and AMD and NVIDIA. China's talents in this area are relatively scarce and scarce in the world. Being able to bring these high-end talents together, he naturally creates a huge threshold.

Odaily Planet Daily: What are your strengths in technical talent? What quantitative indicators can be explained?

Yang Xin: We have all the R&D personnel, and the whole system of R&D talents is the most complete and high-end. Specific can see our prospectus.

Odaily Planet Daily: Are you going to pay attention to some new players in the market? Like I have heard of Yotta, hey, have you heard of these brands?

Yang Xin: We are technically concerned about all kinds of things, but we still talk in kind. PPT mining machines are generally filtered out by us in the early days.

Odaily Planet Daily: I often hear how high the cost of using a tape is to be the threshold for R&D. Is this standard reasonable?

Yang Xin: This is a small threshold. From the point of view of the film, it is a one-time fee, this cost is really high, but it will not let the giants do not play. For example, if ICBC wants to do this thing, it is easy to pay for it. But he does not have this talent. Talent should be the most expensive, and the technical threshold is often much higher than the funding threshold.

The happiest thing is to hear that "the product is sold out"

Odaily Planet Daily: What is your usual working status? What is the happiest time and the toughest time?

Yang Xin: Working status, it is still very hard (laughs).

The happiest time is to see the machine selling. When the market (personnel) came to urge me to say "Hurry up and produce more, sell out." I am very happy at this time. I am recognized for the things I have done, and my mood is still very good.

Odaily Planet Daily: Should it be from the end of 2017 to the first half of 2018?

Yang Xin: There are often. We are still always happy.

When it is difficult, sometimes the mining machine encounters unexpected problems, or it is a long time, long delay, or the phenomenon is more complicated and difficult to locate, which is also very common in research and development. Although everyone will be encouraged at this time, the heart is still relatively heavy. Especially when it comes to problems that can't be solved day and night, it is still very annoying.

Odaily Planet Daily: There will be a period of time to focus on the hard work to make a product?

Yang Xin: Yes, this state is very common here. In order to make this thing as good as possible, many things are a problem and a problem. In terms of performance, it is 1%, 1% to fight for it, and to accumulate to achieve the value you want. There are a lot of efforts from our R&D colleagues.

Odaily Planet Daily: How do you deal with problems that are difficult to tackle, especially when the technology breaks?

Yang Xin: They (the staff of the department) will definitely have it, and they will come to me in the end, and then everyone will work together to find a solution. I also know that this door is difficult in the world. Someone knows these things better than us. Originally, the engineers here are gathered from very good companies. Their own qualities are already top-notch. Coupled with their familiarity with the industry and products, it is difficult for the outside world to understand the industry better than us.

Odaily Planet Daily: Ok, thank you Mr. Yang Xin for sharing today.

Source: Planet Daily

Author: Huang Xue Jiao

We will continue to update Blocking; if you have any questions or suggestions, please contact us!