Lanqi Technology has been investigated by 70 institutions： in 2021, the market share of related products sales of the company is about 2%, which has great market expansion potential (with survey quest

　　Lanqi Technology released the record of investor relations activities on June 6, and the company was investigated by 70 institutions on May 30, 2022. The types of institutions are QFII, insurance companies, others, fund companies, overseas institutions, securities companies and Sunshine private equity institutions.

　　The main contents of investor relations activities are introduced:

　　(I) Brief introduction of the performance in the first quarter of 2021 and 2022. Ms. Fu Xiao, secretary of the board of directors of the company, briefly introduced the performance of the company in 2021 and the first quarter of 2022.

　　[Summary of main operating conditions in 2021] With the company’s Tianjin arrest? A major breakthrough was made in the CPU business, and the first generation memory interface chip and memory module supporting chip of DDR5 were officially mass-produced and shipped in the fourth quarter of 2021. In 2021, the company achieved an operating income of 2.562 billion yuan, a record high in the company’s revenue, with a year-on-year increase of 40.49%, and a net profit of 829 million yuan, down 24.88% year-on-year. In the fourth quarter of 2021, the company’s revenue was 969 million yuan, achieving continuous growth in a single quarter, setting a new record for the company.

　　In the fourth quarter of 2021, the company’s interconnected chip product line revenue was 601 million yuan, a record high for the product line in a single quarter; Tianjin arrest? The annual revenue of the server platform product line was 845 million yuan, a year-on-year increase of 27.5 times.

　　The company insists on technological innovation and attaches great importance to research and development. During the reporting period, the company’s R&D expenditure was 370 million yuan, up 23.33% year-on-year, accounting for 14.44% of operating income.

　　By the end of 2021, the company had 373 R&D technicians, a net increase of 56 compared with the end of 2020, accounting for 71% of the total number of the company, with master’s degree or above accounting for 65%. The company’s R&D technical team has been further expanded and its overall strength has been further strengthened.

　　In 2021, the company obtained 17 authorized invention patents, and the number of authorizations increased by 89% year-on-year; 48 new invention patents were applied, and the number of applications increased by 37% year-on-year; Newly acquired 16 integrated circuit layout designs, up 78% year-on-year; Two new software copyright registrations were obtained.

　　The company and its products have been continuously affirmed by customers and industries, and they have been elected as "Single Champion Demonstration Enterprise of Manufacturing Industry" and "National Technology Innovation Demonstration Enterprise" by the Ministry of Industry and Information Technology, and they have been elected as members of the board of JEDEC (China was selected as the only chip company among the three enterprises), which reflects the company’s position in the industry.

　　[Operating results in the first quarter of 2022] In the first quarter of 2022, the company achieved revenue of 900 million yuan, a year-on-year increase of 201%; The net profit of returning to the mother was 306 million yuan, a year-on-year increase of 128%; Deduction of non-homecoming net profit was 231 million yuan, a year-on-year increase of 211%; The net cash flow from operating activities was 259 million yuan, a year-on-year increase of 101%.

　　As of March 31, 2022, the company’s total assets were 9.4 billion yuan, including net assets of 8.8 billion yuan.

　　In the first quarter of 2022, the company’s interconnected chip product line achieved revenue of 575 million yuan, a year-on-year increase of 94%; Tianjin arrest? The server platform product line achieved revenue of 325 million yuan, a year-on-year increase of 99 times. In the first quarter of 2022, DDR5 related products continued to be shipped, and Tianjin arrested them? The server platform business has developed steadily, and the company’s operation is relatively stable.

　　[R&D progress of major products in 2021 and key points of R&D work in 2022] In addition to products that have been mass-produced in 2021 and products under development, the company has actively made technical reserves in some new technologies and products, laying a solid foundation for the follow-up research and development of new products. In the process of research and development of existing products, the related underlying technologies are constantly precipitated, and the research and development of new products are constantly fed back. On the whole, the company continues to innovate its products around its own technology and market ability, and realizes the positive cycle of "reserve research and development-mass production".

　　The progress and R&D achievements of the company’s main products in 2021 are as follows:

　　(1) interconnection chip product line: the first generation memory interface chip of DDR5 and the supporting chip of memory module are mass-produced, and the second generation memory interface chip of DDR5 and PCIe5.0Retimer engineering samples are completed.

　　(2) Tianjin Capture Server Platform Product Line: In 2021, the third generation Tianjin Capture CPU was launched.

　　(3)AI chip: complete the logic design, system integration and verification of the main subsystems of AI chip, promote the system construction of AI and big data software ecology, and complete the main function verification and performance evaluation of typical application scenarios.

　　In 2022, the company will continue to invest in R&D innovation, and the key points of R&D work are as follows:

　　1. Steadily promote the iterative upgrade and research and development of existing products;

　　(1) interconnection chip product line: complete the research and development of DDR5 second-generation memory interface chip and PCIe5.0Retimer production version, and make preparations for quality certification before mass production.

　　(2) Tianjin arrest? Server platform product line: complete the fourth generation Tianjin arrest? CPU research and development, and realize mass production;

　　(3)AI chip: complete the first generation AI chip engineering sample streaming.

　　2, start a number of new product research and development design work:

　　Including CKD chip, MCR RCD/DB chip and MXC chip, the goal is to complete the first generation of product engineering samples of the above three new products before the end of 2022. At the same time, the company’s new "2022 Restricted Stock Incentive Plan (Draft)" sets the assessment indicators for 2023 as follows: the above three new products have completed the research and development of mass production versions and achieved shipment.

　　(1)CKD chip: DDR5UDIMM and SODIMM for the new generation of desktop and notebook computers.

　　(2)MCR RCD/DB chip: a high-bandwidth memory module for servers and data centers.

　　(3)MXC chip: used for memory expansion and memory pooling of the server.

　　[Product Layout] At present, the company carries out strategic product layout around two areas: high-speed interconnection and high-performance data processing.

　　(1) Interconnected chips belong to "fast chips" and are related to the underlying technology. Based on the technical and market advantages in the field of memory interface chips, the product layout is gradually extended to memory module supporting chips, PCIe and CXL related chips.

　　(2) Data processing chips are "big chips", including Tianjin CPU and AI chips. Is the company developing Tianjin arrest? In the process of CPU, the technical ability in the field of data processing chips is gradually accumulated, and there are more opportunities to meet the needs of server manufacturers and end users. Combined with the company’s technical reserves in DDR and CXL, it laid the foundation for the research and development of AI chips.

　　Generally speaking, the company’s product layout starts with the memory interface chip, and is extended to Tianjin and Shenzhen under the dual drive of technology and market. The server platform, memory module supporting chips, PCIe Retimer chips and AI chips are now further expanded to CKD, MCR RCD/DB and CXL related chips, and we will see the company’s richer product layout in the future. The company’s sub-sectors and products under research have high commercial and technical thresholds. Because the company has certain technical advantages and resource endowments in related sub-sectors, we are confident to participate in the global competition of related products. The company hopes to expand the accessible market scale in the future, consolidate its competitive advantage and lay a solid foundation for long-term healthy and stable development by continuously enriching product categories.

　　(two) the main questions and answers to the exchange.

　　Q: Please look forward to the penetration rate of DDR5 related products.

　　A: In the fourth quarter of 2021, DDR5 memory interface chip and memory module supporting chip have been officially mass-produced and shipped, and will continue to be shipped in the first quarter of 2022. With regard to the penetration rate of DDR5-related products, it is divided into two dimensions: server side, before the server CPU supporting DDR5 is put into use, it is expected that the penetration rate of DDR5 industry will be slowly improved, and the larger penetration rate will climb after the scale of server CPU supporting DDR5 is put into use; On the PC side/laptop side, the desktop platform supporting the first generation of DDR5 was released in the fourth quarter of 2021. At the same time, according to the planning of mainstream CPU manufacturers in the market, the desktop platform supporting the next generation of DDR5 may be launched before the end of this year. These factors will accelerate the upgrade of the desktop from DDR4 to DDR5.

　　Q: How many children does DDR5 memory interface chip have? Compared with DDR 4, is the iteration speed between each offspring of DDR 5 memory interface chip accelerating, and what impact does it have on the company?

　　A: According to the information published by JEDEC, the DDR5 memory interface chip has planned three generations, with support rates of 4,800 mt/s, 5,600 mt/s and 6,400 mt/s respectively, and it is expected that there may be one or two generations in the future. The industry expects that in the early DDR5 generation, the iteration speed between the descendants of memory interface chips may be faster than that of DDR4 generation. On May 9, 2022, the company announced that it was the first in the industry to trial-produce the second generation RCD chip of DDR5. It is expected that this year, in addition to the gradual increase in the first generation of DDR5 products, the second generation of DDR5 memory interface chips will also have a certain amount of sample demand.

　　The acceleration of the iteration speed of DDR5 offspring is of positive significance to the company: first, according to the price law of DDR4 offspring iteration, the initial sales price of the new offspring products is higher than that of the previous offspring due to the upgrading of technology and performance, which can boost the average price of products; Second, as a leader in the segmentation field, the company took the lead in formulating the product standards of the first and second generation memory interface chips of DDR5 in the industry organization JEDEC, which reflected the company’s position in the industry and helped to maintain the company’s leading position in research and development.

　　Q: What is the price law of each offspring of memory interface chip?

　　A: Combined with the price law of DDR4 memory interface chip, the sales price of a certain offspring product gradually decreases after the quantity is added. However, due to the upgrade of technology and performance, the initial sales price of the new generation products is higher than that of the previous generation. Therefore, in the case of continuous iteration, the average sales price of products will remain stable and upward.

　　Q: What is the competition pattern between DDR5 memory interface chip and memory module supporting chip?

　　A: At present, the competition pattern of DDR5 memory interface chips is similar to DDR4 memory interface chips. Only three suppliers in the world can provide the first generation of mass production products, namely, the company, Renesas Electronics and Rambus, and the company’s market share in memory interface chips remains stable. On the supporting chips, the two main suppliers of SPD and TS are Lanqi and Renesas Electronics. PMIC has more competitors, and the competition will be more complicated in the initial stage.

　　Q: What is the reason for the decline in the revenue of interconnected chips in the first quarter of 2022?

　　A: The company’s revenue of interconnected chips decreased slightly in the first quarter of 2022, mainly due to: 1. Quarterly fluctuations. In normal years, the first quarter is off-season compared with the fourth quarter; 2. Due to the special period disturbance, the logistics efficiency of some products was affected by the static management in Shanghai and its surrounding areas in late March, and the shipment was postponed to April. After active coordination, the current logistics efficiency has basically returned to normal.

　　Q: Why did the gross profit margin of interconnected chips decrease in the first quarter of 2022?

　　A: In the first quarter of 2022, the gross profit margin of interconnected chips decreased slightly from the previous quarter, mainly because: DDR5 memory module supporting chips were jointly developed by the company and its partners, and the gross profit margin was lower than that of memory interface chips. With the increase of its revenue share, the average gross profit margin of interconnected chip product lines declined.

　　Q: Does the company have room to improve the future market space of CPU business? What are the direct customers and downstream applications of this business?

　　A: In view of Tianjin CPU, because the market space is very large, the company’s primary goal in the initial stage of products is to pursue quantity, enter the market as soon as possible, strive for more market share and gain more customers’ recognition. Because of the large value of this product, once it is on the market, the absolute amount of gross profit will be considerable. At the same time, after accumulating a certain number of customers and reaching a certain income scale, the company will develop new functions to enhance the added value of products in combination with the needs of customers, and actively optimize costs, thus ultimately improving the gross profit margin of this product line.

　　China’s server market maintained a high-speed growth trend. According to IDC’s research report, in 2021, the shipment and sales of China’s server market were 3.911 million units and 25.09 billion US dollars, respectively, up by 8.4% and 12.7% year-on-year. The introduction of strategies such as "Digital China" and "Counting East and Computing West" is expected to bring great impetus to the development of China server market in the next few years. At the same time, with the continuous strengthening of the competitive strength of domestic brands, domestic brands’ server-related products will enjoy more growth dividends in China server market.

　　In 2021, the sales volume of related products of Tianjin Capture Server Platform exceeded 150,000 pieces. If a server is equipped with two CPUs on average, the market share of related products of the company will be about 2% in 2021, which has great market expansion potential.

　　The direct customers of the company’s CPU are mainly OEM or ODM manufacturers of servers, and the larger downstream applications are finance, transportation, government affairs, energy, data centers and other fields.

　　Q: The company’s AI chips are strategic products in the future. Please introduce the technical architecture, application scenarios and market prospects of AI in detail.

　　A: 1. Technical architecture.

　　The AI chip solution under research by the company consists of AI chips and other related hardware and corresponding adaptation software, and adopts a near memory computing architecture, which is mainly used to solve the CPU bandwidth, performance bottleneck and GPU memory capacity bottleneck problems in reasoning application scenarios of AI computing under big data throughput, and to provide customers with low latency and high efficiency AI computing solutions.

　　AI chip is the core hardware of the above solution, which is mainly composed of AI computing subsystem, CXL controller, DDR memory controller and other modules. The chip is designed specifically for the application of AI in big data scenarios, and integrates AI high-performance computing, heterogeneous computing, CXL high-speed interface technology, DDR memory control technology and other related technologies. It has efficient hardware acceleration functions such as searching and sorting large-capacity data, and has functions such as data compression and data encryption and decryption.

　　At the same time, the company’s AI chip solution will support a perfect AI software ecosystem, and can deeply optimize all kinds of AI algorithms and models in a targeted way, and can support all kinds of mainstream neural network models in the industry, such as visual algorithms, natural language processing and recommendation systems, which is conducive to the subsequent software and hardware ecological construction and marketing work.

　　2. Typical application scenarios of AI chips in the future The typical application scenarios of AI chips under research by the company in the future are as follows:

　　(1) Recommendation system under big data throughput in the Internet field. At present, the conventional scheme in the industry is to hand over the two main steps of "Embedding" and "Embedding Search" in the recommendation system to different platforms for processing, and the GPU, FPGA or ASIC chip with high computing power is responsible for the "Embedding" part, and the CPU+ big data system deploys the "Embedding Search" part. This step is divided, resulting in a lot of data exchange, and due to the hardware, The goal of the company’s AI chip is to integrate the above two steps and balance the computing power and memory capacity at the same time, so that the computing resources and memory can be used efficiently and the efficiency bottleneck of the system can be solved.

　　(2) Biomedical/medical big picture stream processing in medical field. At present, the conventional scheme in the industry is to cut large pictures in CPU, and the sub-images obtained by cutting are transmitted to GPU for AI processing through PCIe interface. Through many interactions, a big picture is finally processed, which is also limited by the interface bandwidth and memory between them.

　　The company’s AI chip can greatly increase the memory capacity and reduce or even eliminate the need for image cutting. At the same time, CXL interface can make full use of cache performance and directly access the DDR memory of the near-memory computing module, thus improving the efficiency of the interface.

　　(3) Big data application scenarios in the field of artificial intelligence Internet of Things.

　　Generally speaking, the goal of the company’s AI chip solution is to provide customers with more efficient and cost-effective solutions compared with traditional solutions in similar application scenarios.

　　3. Market prospect: In recent years, the market scale of AI servers has been expanding. IDC predicts that the global market scale of AI servers will reach $27.7 billion in 2025, with a compound annual growth rate of 20.3%. According to the application field, AI chip can be used in two scenarios: training and reasoning. According to the prediction of relevant institutions, the global cloud AI chip market will reach a compound growth rate of 28% in 2020-2025 and a scale of 26.1 billion dollars in 2025: the market scale of cloud training chips will reach 12.9 billion dollars in 2025, and the compound growth rate will be 25% in 2020-2025; In 2025, the cloud inference chip will reach the market scale of 13.2 billion US dollars, and the compound growth rate in 2020-2025 will be 32%, which is higher than the overall growth rate of cloud AI chips. The company’s AI chip is facing the reasoning application scenario under the application of big data throughput. Because it is a new product, the current market scale is difficult to predict, but the value of AI chip is large. This is a very important product in the strategic layout of data processing chips, and I hope this product can become a new revenue growth point for the company in the future.

　　Q: The newly-developed CKD chip is mentioned in the company’s annual report. Please tell us in detail the function, application and future market space of this chip.

　　Answer: CKD chip is a high-speed Clock Driver defined by JEDEC organization, which can buffer and re-drive the clock signals of the next generation UDIMM and SODIMM (mainly used for desktops and laptops) of DDR5, thus improving the signal integrity and reliability of the clock signals and providing a higher-speed memory solution for desktops and laptops.

　　It is expected that in the middle of DDR5, "CKD+SPD+PMIC" will become the standard of UDIMM and SODIMM for desktops and laptops, and the direct customers will be memory module manufacturers. If CKD chips are fully infiltrated, their demand will be positively related to the global demand for UDIMM and SODIMM for desktops and laptops.

　　According to IDC data, the global shipments of desktops and laptops in 2021 were 349 million units, an increase of 14.8% compared with 2020.

　　At present, JEDEC is still formulating the product standard of CKD chip, and the company has participated in the standard formulation and laid out the research and development. The research and development goal of the company’s CKD chip before the end of 2022 is to complete the flow of engineering samples. At the same time, the company’s new "2022 Restricted Stock Incentive Plan (Draft)" sets one of the assessment indicators for 2023: CKD chips complete the research and development of mass production versions and realize shipment.

　　Q: official website recently released the world’s first MXC chip. Please tell us more about its function, application and future market space.

　　A: On May 6, 2022, the company released the world’s first CXL (Compute Express Link) memory expansion controller chip (MXC). The full name of MXC is Memory Expander Controller, and its Chinese name is Memory Expander Controller. The MXC chip is specially designed for memory AIC expansion card, backplane and EDSFF memory module, which can greatly expand the memory capacity and bandwidth to meet the growing demand of data-intensive applications such as high-performance computing and artificial intelligence. MXC chip is a CXL DRAM memory controller, which belongs to the third device type defined by CXL protocol. The chip supports JEDEC DDR4 and DDR5 standards, and also conforms to CXL2.0 specification, supporting the speed of PCIe 5.0. This chip can provide a high-speed interconnection solution with high bandwidth and low latency for CPU and devices based on CXL protocol, thus realizing memory sharing between CPU and CXL devices, greatly improving system performance, and significantly reducing software stack complexity and total cost of ownership (TCO) of data center.

　　MXC chips are mainly used in memory expansion and pooling application scenarios of big data, AI and cloud services. Before CXL interconnection technology, the bottleneck of memory expansion lies in the limitation of the number of CPUs and memory slots and the limitation of the memory capacity corresponding to each CPU. The difficulty of memory pooling lies in the fact that memory can only be accessed by CPU and cannot be shared by other devices.

　　With CXL interconnection technology, the memory expansion card can be composed of MXC chips and DIMM particles to realize the function of memory expansion or pooling, which is beneficial to the flexible allocation of memory space in data centers and effectively reduces the total cost of ownership.

　　The market demand of MXC chip depends on the market demand for memory expansion and memory pooling. Because it can make the memory configuration of data center more flexible and help customers reduce the total cost of ownership, the company is very optimistic about the future prospect of this product.

　　On May 10th, 2022, Samsung Electronics announced the development of its first 512GB CXL DRAM. The CXL internal control extended controller chip (MXC) of this memory module was provided by the company.

　　One of the assessment indicators in 2023 set by the company’s newly launched Restricted Stock Incentive Plan (Draft) in 2022 is: MXC chips complete the research and development of mass production versions and realize shipment. Lanqi will continue to maintain close cooperation with the leading memory manufacturers in the industry to jointly build an ecosystem based on CXL memory expansion controller and constantly update its products.

　　It is generally believed in the industry that CXL technology has great market potential in the future. According to the relevant materials of the investor briefing held by Micron Technology in May this year, it is believed that the market scale of CXL-related products will reach 2 billion US dollars by 2025 and further exceed 20 billion US dollars by 2030, driven by the rapid development of heterogeneous computing.

　　Q: The newly-developed MCR RCD/DB chip is mentioned in the company’s annual report. Please tell us the function, application and future market space of this product in detail.

　　A: MCR RCD/DB chip is used for MCR memory module, which is a higher bandwidth memory module. With the increasing number of AI processing in the cloud, the demand of high throughput, low latency and high density has spawned the demand for higher bandwidth, faster speed and higher capacity memory module. In order to meet this demand, JEDEC organization is currently formulating relevant technical standards for server MCR memory module.

　　The MCR memory module adopts the "1+10" infrastructure of LRDIMM. Compared with ordinary LRDIMM, the MCR memory module can access two arrays on the memory module at the same time, providing double bandwidth. The first generation of products can support up to 8800MT/s, and it is expected that there will be two or three generations of products with higher speed in the DDR5 generation. The memory interface chips needed for the high bandwidth memory module of the server are MCR RCD chip and MCR DB chip. Compared with ordinary RCD chip and DB chip, the design is more complicated and the speed is higher.

　　At present, JEDEC is still formulating the standard of MCR RCD/DB chip, and the company has participated in the standard formulation and has laid out research and development. Compared with ordinary RCD/DB chips, MCR RCD/DB chips have higher technical requirements, more difficult research and development, and longer research and development cycle, so it is difficult for manufacturers who have not developed RCD/DB to break through in a short time. The research and development goal of the company’s MCR RCD/DB chip before the end of 2022 is to complete the flow of engineering samples. At the same time, the company’s new "Restricted Stock Incentive Plan for 2022 (Draft)" sets one of the assessment indicators in 2023 as follows: MCR RCD/DB chip completes the research and development of mass production version and realizes shipment.

　　Q: What is the mass production and shipment of PCIe4.0Retimer chip in 2021, the future market situation and the research and development progress of PCIe5.0Retimer chip?

　　A: The company’s PCIe4.0Retimer chip achieved sales revenue of 12.2 million RMB in 2021, achieving a substantial breakthrough from zero to one.

　　From an industry perspective, in the short term, the PCIe4.0 ecosystem has just started, and the overall scale of the industry is not large; Looking into the future, it is generally believed in the industry that the demand for PCIe4.0-related equipment will gradually increase this year. At the same time, the mainstream server CPU supporting PCIe5.0 will be listed this year, and the demand for PCIe5.0-related equipment will enter the start-up period, which will also drive the demand for PCIe Retimer chips, and will drive the improvement of PCIe5.0 ecosystem and the commercialization process of related products. Therefore, the whole industry is still very optimistic about the PCIe Retimer chip market in the medium term. In January 2022, PCI-SIG officially released the final (1.0) specification of PCI Express6.0. Compared with PCIe5.0, the transmission speed doubled again, from 32GT/s to 64 GT/s.

　　In 2021, the company’s PCIe5.0Retimer chip has completed the slide of engineering samples, device and system-level test and evaluation, and began to send samples to customers and partners for interoperability testing. Before the end of 2022, the company’s PCIe5.0Retimer chip aims to complete the research and development of mass production version.

　　Q: The company’s R&D direction is the application field of SerDes high-speed interface. Is it the direction of the company’s future product line expansion?

　　A: SerDes is short for SERializer /DESerializer.

　　It is a mainstream serial communication technology of time division multiplexing (TDM) and point-to-point (P2P). That is, at the sending end, multiple low-speed parallel signals are converted into high-speed serial signals, which pass through the transmission medium (optical cable or copper wire), and finally at the receiving end, the high-speed serial signals are converted into low-speed parallel signals again. As an important underlying technology, SerDes is usually used as the physical (PHY) layer of some important protocols (such as PCIe, USB, Ethernet, etc.) and is widely used in high-speed interconnection in the fields of servers, automotive electronics, communications, etc.

　　Because SerDes is an important basic technology in the field of high-speed interconnection, which can be applied to PCIe Retimer and many other products, the company attaches great importance to this technology, and will also make some new product layouts based on the extension of technology. The specific new product layout strategy will be gradually developed in combination with the company’s technical advantages or market advantages.

　　Q: What is the strategic layout of the company’s future products?

　　A: The company will focus on scientific and technological innovation in the field of integrated circuit design, constantly meet customers’ demand for high-performance chips around the fields of cloud computing and artificial intelligence, realize the leap-forward development of the enterprise through continuous accumulation, create good returns for shareholders and contribute beneficial value to society.

　　The development goal of the company in the next three years is to enhance the company’s market position and influence in sub-sectors through continuous R&D and innovation, and at the same time open up new business growth points.

　　Among them:

　　1. In the field of interconnected chips, on the one hand, the company will continue to consolidate its market leading position in DDR5 related businesses, and continue to invest in iterative research and development of DDR5 related chips and new generation products of PCIe Retimer chips to further improve product layout; On the other hand, the company will continue to pay attention to other market opportunities in the field of interconnected chips and make strategic layout in a timely manner. In the annual report of 2021, the company has disclosed the research and development plans of three new products, namely CKD chip, MCR RCD/DB chip and MXC chip. They all belong to the field of interconnected chips and involve cutting-edge technologies in the industry. The company will steadily promote the research and development of new products as planned.

　　2. In the field of data center business, continue to upgrade Tianjin? Server CPU and its platform provide high-performance, high-security, high-reliability CPU, mixed secure memory module and other products for data centers, and continuously increase market share.

　　3. In the field of artificial intelligence chips, the company will focus on customer needs, tap potential business opportunities, develop competitive chip solutions, and provide new business growth points for the company’s sustainable development. The artificial intelligence chip currently being developed by the company adopts the near memory computing architecture, which is mainly used to solve the CPU bandwidth, performance bottleneck and GPU memory capacity bottleneck problems in the reasoning application scenario of AI computing under big data throughput, and provide customers with low latency and high efficiency AI computing solutions.

　　4. The company will continue to pay attention to the development of the industry and use its own resources to find suitable industrial chain investment and M&A opportunities.

　　Q: What is the impact of this epidemic in Shanghai on the company’s production and operation?

　　A: Due to the need of epidemic prevention and control, some employees of the company have been working at home recently, and some employees have been arranged to be on duty at the company site. After active coordination, the logistics efficiency of the company’s related products around Shanghai has returned to normal. Up to now, the company’s operation is generally normal.

　　Q: Does the current tight supply chain in the semiconductor industry affect the company?

　　A: From the second half of 2020, the company predicted in advance that the production capacity of the whole industry would be relatively tight by 2021, so it made some plans in advance. Up to now, the production capacity of the company’s main products is basically guaranteed, but it is still possible to be nervous about some products or models.

　　Q: How many R&D personnel are there in the company? What is the company’s R&D personnel expansion plan? What is the distribution of personnel in different product lines, and what are the main aspects of the increase of R&D personnel?

　　A: Since the company went public, the number of employees has increased steadily, from 255 at the end of 2018 to 334 at the end of 2019 and 457 at the end of 2020 to 527 at the end of 2021, among which the proportion of R&D technicians has increased steadily, reaching 71% at the end of 2021, with 373 employees. With the development of the company’s business and continuous R&D investment, the company’s R&D technicians will continue to maintain a reasonable growth. The characteristics, technical reserves and R&D stage of each R&D project will be different, and the company will arrange corresponding R&D personnel and investment according to the needs of different R&D projects. Generally speaking, the increase of the company’s R&D personnel is mainly concentrated on new products, including AI chips and PCIe Retimer chips.

　　Q: Does the company consider investing in mergers and acquisitions to accelerate its development?

　　A: On the one hand, the company will achieve high-quality endogenous growth by strengthening R&D, on the other hand, it will continue to pay attention to the development trends of the industry and use its own resources to find suitable industrial chain investment and M&A opportunities;

　　The main business of Lanqi Technology Co., Ltd. is to provide chip-based solutions for cloud computing and artificial intelligence; The company’s main products are memory interface chip, Tianjin server platform and consumer electronics chip; In 2018, the company was awarded the 20th China Patent Excellence Award by China National Intellectual Property Administration. In 2018, the company’s product "the second generation DDR4 memory buffer controller chip" won the "China Core" annual major innovation breakthrough product award; In November, 2018, the "Dynamic Security Monitoring Technology" adopted by the CPU of Tianjin Capture Server and its platform was awarded the "World Internet Leading Scientific and Technological Achievements" in the 5th world internet conference.

　　Details of participating institutions are as follows: