Image source @ Visual China
Wen 丨 core lithium words
Looking back at the photovoltaic industry in 2021, silicon material companies have become the biggest winners in the industrial chain, and the price increase caused by tight supply and demand has made silicon material companies gain a lot.
The rise in prices often means the enhancement of the right to speak in the industrial chain, and in recent times, the price of key materials has skyrocketed in the chip industry chain.
At the beginning of November last year, there was a sudden shortage of "photomask" supply in the chip industry chain, especially in the field of automotive chips, this shortage was more obvious, not only the delivery time was greatly extended, but also the price increased generally, and some high-end products even directly increased the price by 25%.
This can't help but make the "light mask" unpopular track regain our attention.
China's chip industry started late, and is currently facing a more serious card neck problem, such as lithography machine, EDA software and other core chip manufacturing links We have obvious industrial gaps.
In the field of chip materials, we are also restricted by Japanese companies, photoresist, CMP polishing materials, wet chemicals and other chip materials are still mostly imported from overseas, "photomask" is one of them.
From the perspective of scale, "photomask" only accounts for about 13% of the total cost of the chip, and its value is much lower than that of silicon wafers, which account for 38%, and the attention is far from silicon wafers.
However, the value of a small "photomask" is extraordinary, it is not only one of the indispensable core materials in chip manufacturing, but also its quality directly determines the final performance of the chip.
In the context of the gradual transfer of global chip production capacity to China, the independent replacement rate of mainland chip materials is extremely low, especially in the overseas epidemic, chip materials have the risk of supply interruption at any time, and the early completion of independent substitution has become the consensus expectation of China's chip industry.
The crazy rise in the price of "light masks" has sent a signal to the market, and the unpopular track also has fanatical dreams.
01, Japan choked the throat of global semiconductors
Chip manufacturing is an extremely complex process, and as the process progresses, the steps of chip manufacturing are increasing.
But no matter how the steps of the chip increase, the first step in chip manufacturing must be lithography, that is, the circuit group designed in advance is engraved into the wafer through the lithography machine.
In the entire lithography process, in addition to the indispensable lithography machine, the photomask is also impossible or absent. It is like the negative of the camera, and the lithography machine is equivalent to an exposure device, and the two work together to produce the final photo.
Further, it is not the lithography machine that determines the circuit pattern, but the structure of the photomask, so the photomask is like the throat of the chip industry chain, and only through it can the function of the chip be finally realized.
The accuracy and detail of the photomask is directly related to the quality of the semiconductor chip, and the photomask is not like the chip is a pipeline operation, each mask requires a special design, so this industry presents a high degree of customization characteristics.
According to the classification, photomask factories can be divided into two categories: fabs themselves supporting factories and independent third-party mask manufacturers. Due to the technical confidentiality of each wafer manufacturing plant in chip manufacturing and design, fabs often independently develop mask versions of advanced processes below 45nm, while mask versions of mature processes above 45nm are handed over to third-party mask factories for research and development.
According to SEMI's data, in the global chip mask plate market in 2019, 65% of the market share was occupied by the fab's own mask plate factory, and the remaining 35% of the share was divided by independent third-party mask factories.
At present, the global core third-party semiconductor photomask production capacity is mainly concentrated in Japan, the global Japanese toppan and Japanese DNP companies occupy 32% and 27% of the global third-party market share respectively, the United States Photononics occupies 23% of the third-party market share, and the remaining share is divided by the waist tail company.
Because fabs often choose to develop masks for advanced processes on their own, many investors mistakenly believe that photomasks are an industry without too high technology content. But in fact, on the contrary, the production of masks is equivalent to a small lithography process, and enterprises with independent R&D and production capabilities of photomasks are actually based on technology.
Japan has more than 60% of the world's third-party photomask production capacity, and it already has the ability to choke the throat of the semiconductor industry. Although there is no independent third-party photomask factory to help, chip factories still have independent mask development capabilities, but this will undoubtedly greatly reduce the development speed of the semiconductor industry and affect the iterative efficiency of the entire industry.
Based on this, if China's chip industry wants to catch up with the global development speed, then the independent replacement of photomask is imperative.
02, the difficult start of domestic photomasks
Photomasks are not standardized products like wafers and often need to be customized according to the needs of the user. Therefore, the price of the photomask actually depends on the relationship between supply and demand.
Third-party mask manufacturers do not carry out simple processing, but need to complete up to 14 steps to finally deliver the order, including many core steps such as lithography, development, etching, cleaning, and detection.
That is to say, photomask factories not only need to purchase these semiconductor devices, but also skillfully use them.
Overall, the production of photomasks is a consideration of the comprehensive capabilities of the company's semiconductors, although it does not require the completion of a large number of tape-out projects, but also needs to have the ability to produce independently from beginning to end.
Compared with Japanese manufacturers, China's third-party masking companies are lagging behind in all aspects and are still in the stage of technology accumulation.
Due to the late start of the semiconductor industry in mainland China as a whole, the scale of related enterprises is small, the profitability is poor, it is difficult to obtain effective financing channels, and financing difficulties once became an obstacle to the development of the semiconductor industry in mainland China.
However, after the launch of the science and technology innovation board, the original financing bottleneck was broken, and many semiconductor industry chain companies with weak profitability were also listed, and gradually expanded the volume, and China's third-party photomask enterprises were one of the beneficiaries after the launch of the science and technology innovation board.
In 2019, Qingyi Optoelectronics landed on the Science and Technology Innovation Board, becoming the first photomask company to land on the capital market; coincidentally, Luwei Optoelectronics also submitted an IPO application in 2021, which is expected to become the second photomask company to complete the listing.
Although it has gradually landed on the capital market, there is still a significant gap between both revenue scale and technical strength. Since photomasks are not only essential materials for semiconductors, but also for LCD screen production, Chinese photomask companies generally choose the way of "screen tape core" and put chip business in a secondary position.
According to the announcement data, the total revenue of Qingyi Optoelectronics in the first half of 2021 was 228 million yuan, of which the semiconductor business revenue was 39.7 million yuan, accounting for about 17.4% of the total revenue. Luwei Optoelectronics' total revenue in 2020 is 402 million yuan, of which the semiconductor business revenue is 86.63 million yuan, accounting for about 21.5% of the total revenue.
It can also be seen from the company's name that the definition of photomask companies is still more inclined to the display industry.
In addition to the poor strategic focus, there are still industrial chain faults in China's photomask industry. The mask version of the key upstream material substrate relies on imports from Japan. This means that even if Chinese photomask companies grab orders from overseas giants, upstream materials still need to be imported from Japan, adding a lot of helplessness.
The core strength is poor, the industrial chain is faulty, which is the actual dilemma facing China's photomask enterprises. In this context, how should we achieve a breakthrough?
03, the missing upstream industry chain
The lack of industrial chain makes China's photomask enterprises always subject to people, but what is more regrettable is that the mainland photomask industry chain is only a little bit away from being opened.
The upstream material of the semiconductor photomask is the processed synthetic quartz plate, but in fact, there is no shortage of synthetic quartz production capacity in the mainland, and Phillips is the leading company in China's synthetic quartz.
At the 2019 performance briefing meeting of Phillips, Zheng Wei, secretary of the company's board of directors, said at the event that Qingyi Optoelectronics is the end customer of Phillips photomask substrates. In the prospectus of Qingyi Optoelectronics, it is difficult for us to find the name of Phillips.
Could it be that Zheng Wei is lying? Of course not, the reason why Qingyi Optoelectronics did not choose to purchase directly from Phillips is because of the lack of intermediate processing links. After the mask foundation leaves the factory, it actually needs to go through several finishing links such as grinding, polishing, chrome plating, photoresist coating, etc., and the processed fine mask substrate is delivered to the photomask manufacturer.
According to the prospectus of Luwei Optoelectronics, at present, there are two companies in the world with mask substrate finishing, Japan HOYA and South Korea LG-IT, and other companies do not have masking and chrome plating processing capabilities, which forces us to choose only imported finished substrates.
That is to say, the lack of mask substrate finishing links has led to the passive position of mainland photomask enterprises. If you want to achieve the independent replacement of photomasks, the premise must be to open up the entire upstream industrial chain, otherwise it is meaningless.
For the lack of industrial chain, whether it is the upstream Phillips or the downstream Qingyi Optoelectronics, in fact, they are well aware. So they have already begun to try to expand into the middle reaches of the industry.
Phillips invested in the establishment of a subsidiary in Hefei, Guangwei Optoelectronics, involved in TFT-LCD and semiconductor photomask template finishing projects, but did not involve the field of gluing and chrome plating; Qingyi Optoelectronics also laid out production lines in Hefei and carried out gluing research and development. In this way, chrome plating is the biggest gap in the technical link.
Although from the current point of view, there is still a technical gap in the mask substrate, the good point is that the upstream and downstream of the industry have begun to try to open up the entire industrial chain.
Especially after the launch of the science and technology innovation board, more and more capital can be invested in the construction of the semiconductor industry, which is expected to accelerate the integration of the industrial chain and make up for the gap in technology.
In the first half of this year, the growth rate of Qingyi Optoelectronics' semiconductor business is much higher than that of liquid crystal business, and semiconductor masks have become the core of Qingyi Optoelectronics' revenue growth, and are expected to benefit from the development of the semiconductor industry and continue to grow.
Looking to the future, once the entire finishing mask substrate industry chain runs through, then the added value of products in the upstream and downstream of the mainland industrial chain is expected to be greatly improved, so as to truly achieve domestic substitution, and the market share of mask companies will be more meaningful.