Continuous losses, market value reduced by 87 billion, AI chips can not support the Cambrian hundreds of billions of market value

Text/Yang Jianyong

A few days ago, Cambrian issued a performance pre-loss announcement, and the full year of 2021 is expected to lose 765 million yuan to 935 million yuan, compared with 2020, the loss expanded by 76% to 115%. Cambrian, which has been expanding in the scale of losses, has continued to slump in the capital market, and its market value has shrunk by more than 80 billion yuan compared with historical highs. It should be known that as an AI chip unicorn, it landed on the science and technology innovation board in July 2020, attracted countless eyes on the trading day, was favored by capital, and rose sharply, once breaking through the 100 billion mark during the session, and in the following trading days, the highest market value was as high as 119.1 billion yuan.

Once the market value was as high as 100 billion, marveling at the power of capital, it was also a cambrian highlight moment. Unfortunately, the current market value of 31.7 billion yuan, compared with the previous high market value has fallen by 73%, the market value has decreased by more than 87 billion yuan. In the author's view, continuous losses are difficult to support the market value of 100 billion, chips as a technology-intensive industry, need to invest in a huge amount of research and development funds, but the high research and development expenditure makes the hematopoietic ability poor, profitability is worrying, the past four years the cumulative loss of more than 2 billion yuan, and the loss is expanding day by day.

In response to the continuous loss of the situation, Cambrian pointed out in the previous financial report: mainly in order to further promote the "cloud edge" industrial layout, expand the product line, increase R & D investment and talent introduction. In terms of R&D investment, R&D expenses in 2021 were 1.044 billion yuan to 1.276 billion yuan, an increase of 35.89% to 66% year-on-year.

In the performance forecast, Cambrian said that the significant increase in research and development expenses is mainly due to the continuous maintenance of the core competitive advantage in the industry, in the case of the shortage of talents in the chip industry, actively introduce outstanding talents, maintain the stability of the company's research and development team, and the salary and expenses of research and development personnel have increased to a certain extent. Of course, smart chip research and development requires a lot of expenses, and for some time to come, there will be continuous losses and potential risks, and Cambrian cannot guarantee profitability in the next few years.

At the same time, in the face of competition from chip manufacturers such as Intel, Nvidia, and ARM, and competition with these chip manufacturers, there is a big gap in cambrian overall scale, financial strength, and research and development reserves. In particular, NVIDIA is the global leader in artificial intelligence chips, with absolute advantages in the global market, including Google, Microsoft and other global technology giants rely on their GPUs to train AI, making it a ride in the global artificial intelligence market, becoming the world's largest chip manufacturer, with a current market value of $627.7 billion.

Fortunately, in terms of revenue, Cambrian has maintained a strong growth trend. According to the performance forecast, the annual revenue scale in 2021 is expected to be 670 million yuan to 7.5 trillion yuan, an increase of 45.99% to 63.42% over the same period of the previous year. Mainly the MLU220 chip and acceleration card of the edge product line have been implemented in many head AI companies, and the large-scale sales of nearly one million pieces have been achieved in the reporting period. In addition, the revenue of intelligent computing cluster systems has increased compared with 2020. Subsequently, the scale of revenue in 2021 increased significantly compared with the same period of the previous year.

As a company that has focused on the research and development and technological innovation of artificial intelligence chip products since its inception, it is committed to creating the core processor chip in the field of artificial intelligence. At present, the main product lines include cloud product lines, edge product lines, processor IP licensing and software. From cloud reasoning Siyuan 270, edge reasoning Siyuan 220, cloud training Siyuan 290, to the newly released Tuixun Siyuan 370, is a chip design company with software and hardware full-stack system capabilities, which can provide customers with different sizes, all categories and multiple scenarios of products and unified platform-based basic system software to meet the differentiated needs of customers.

Among them, Siyuan 370 is mainly oriented to high-end training scenarios, and the Siyuan 370 series acceleration card has carried out in-depth application adaptation with domestic mainstream Internet manufacturers, and the performance performance in speech, vision and other scenarios is relatively outstanding. Edge intelligent chip is a high-performance chip for edge intelligent processing low latency, low power consumption and small size requirements, can land a wide range of application scenarios, support intelligent data analysis and modeling under the edge computing scenario, vision, speech, natural language processing and other diversified artificial intelligence applications, its number of customers is larger, in the field of artificial intelligence edge computing, vehicle-road collaboration, power, energy, education and other industries of the head company carried out in-depth technical cooperation, in 2021 in the state of continuous promotion and shipment.

In the field of artificial intelligence edge computing, Cambrian and a number of head application companies have achieved product introduction, and some enterprises have completed the adaptation work and realized large-scale shipments, through artificial intelligence empowerment, to promote the intelligent upgrading and transformation of the industry. For example, in the field of smart transportation, Cambrian and industry leading partners to build vehicle-road collaboration, smart iron construction and other solutions, with the help of Cambrian diversified intelligent computing equipment, effectively improve the cloud and edge computing power, help customers enhance the reliability of smart tracks and transportation systems, and effectively promote the intelligent upgrading of the transportation industry.

Edge computing is a new computing paradigm that has emerged in recent years, equipped with moderate computing power on devices between the terminal and the cloud. On the one hand, it can effectively make up for the disadvantages of insufficient computing power of terminal equipment, and on the other hand, it can alleviate potential problems such as data privacy, bandwidth and latency in cloud computing scenarios, mainly for customers with high requirements on data security and system timeliness. At the edge, Cambrian seizes the strategic opportunity period of smart technology to begin to enter various traditional industries, with a view to making a major breakthrough in market penetration through higher quality products and services. In the 2021 interim report, the operating income contributed by the edge business accounted for 60.7%, while the operating income contributed by the cloud business accounted for 32.50%.

In addition, Cambrian has entered the field of in-vehicle smart chips. With the in-depth expansion of intelligent driving application scenarios, high-level intelligent driving will inevitably produce higher artificial intelligence computing needs, Cambrian subsidiary Xingge Technology is designing and developing on-board intelligent chips for high-level intelligent driving application scenarios, superimposed on the basis of existing chip technology components to design chips that meet the requirements of vehicle regulations, to help build a unified intelligent ecology of "cloud-edge end cars".

Finally, information technology has developed at an alarming speed, and a wide range of intelligent devices have access to the network, so that information technology has moved towards the Internet of Everything, and driven by new technologies such as artificial intelligence, the Internet of Everything has then changed to the intelligence of all things. In the era of Internet of Everything, the Internet of Everything formed by tens of billions of smart devices, and the strong demand for chips for massive Internet of Things smart devices, from the cloud to the network to the edge, releasing new growth points for chip manufacturers.

Yang Jianyong, a contributor to Forbes China and winner of the NetEase 2020 Most Influential Award, is committed to in-depth interpretation of cutting-edge technologies such as the Internet of Things, cloud services and artificial intelligence, and his views and research strategies have been cited by many authoritative media and well-known enterprises.