close
df3768f2-c479-11e9-8596-aaaa00151f43.jpg

Editor’s note: This article is from the micro-channel public number “grind stone Business Review” (ID: libusiness), Author: Dongmei, 36 krypton release authorized.On August 16, Nvidia announced its fiscal year 2020, which is the second quarter of 2019. The report shows that NVIDIA’s revenue for this quarter was $2.579 billion, a decrease of 17% compared with $3.123 billion in the same period last year;Net profit was $552 million, a 50% decrease from the $1.11 billion in the same period last year.This sharp decline also appeared in Nvidia’s share price in January this year.On January 28th, Nvidia’s share price plunged nearly 14%, and the market value evaporated by 14 billion US dollars overnight. It was described as losing more than 10 million NVIDIA new graphics cards RTX 2080Ti, which caused heavy losses.Since the highest price of $292 in October 2018, it has fallen by more than 50%. Founded in 1993, Nvidia is a manufacturer of GPU chip graphics cards. Since the invention of GPU in 1999, it has redefined modern computer graphics technology.The design and manufacture of graphics cards has greatly promoted the development of the PC game market.Since 2014, the company has been transforming into artificial intelligence, applying GPU graphics computing model to the field of deep learning, and customizing the optimized software and hardware system for deep learning, and has achieved some innovations.What is the difference between Intel and Intel in the chip field, with AMD in the graphics card world, and the star company that Wall Street has won the altar?In the third quarter of 2018, NVIDIA’s revenue has increased for five consecutive years.Especially since 2016, with the development of artificial intelligence and blockchain technology, thanks to the strong demand for products, NVIDIA has gained a lot in the primary and secondary markets and has become a dazzling star.Among them, in the whole year of 2017, Nvidia’s revenue increased by nearly 40%, and profit growth nearly doubled.In the third quarter of 2018, Nvidia achieved $3.181 billion in revenue, an increase of 21% compared to $2.636 billion in the same period last year; net profit was $1.23 billion, an increase from the $838 million in the same period last year.47%.However, when the flowers are blooming and the fire is burning, the situation is abrupt.Since the third quarter of 2018 did not meet Wall Street expectations, Nvidia’s share price plummeted afterwards, after which it lowered its revenue forecast for the fourth quarter of 2018 from the initial $2.7 billion to $2.2 billion, a drop of nearly 19%.This sudden and sudden brake shocked a lot of investors.Subsequently, Nvidia CEO Huang Renxun wrote a letter to the shareholders. In the letter, he summarized the reasons for this mutation as follows: First, the demand for cryptocurrence decreased, resulting in a large inventory; second, NVIDIA high-end graphics cards are too expensive, users stillOn the other hand, the demand for GPUs in data center users is still uncertain; the fourth is the global economic downturn, especially the weak demand in the Chinese market, and customers are cautious when placing orders.In the first quarter of 2019, Nvidia’s revenue declined by 31%, especially in the game segment, which was nearly 40%. Net income in the first quarter fell by nearly $1.29 billion from last year.60%, fell to 543 million US dollars.In the earnings report, this decline was attributed to the “loss” of 289 million US dollars in mining products revenue.Since the middle of 2018, the development of blockchain technology has triggered a boom in mining, with bitcoin being the most expensive and the highest price of more than 20,000 US dollars.Mining requires a lot of graphics resources, but for a while, graphics mining is threatened by FPGA and ASIC mining.The large-scale integration of ASIC mining and the decentralization advocated by the blockchain itself is unfavorable to the long-term development of blockchain technology. Therefore, ASIC was later resisted by the mining market, and the status of graphics card mining was consolidated.The mining graphics card is most popular with Nvidia. With the influx of tidal miners, NVIDIA graphics cards have been sold out and prices have risen sharply.However, the enthusiasm of the cryptocurrency is also fast, and when the craze fades, the market is bleak. In the first month or two after December 2018, the price of bitcoin has fallen by more than 70%, with more than 600,000 miners.shut down.The retreat of the mining boom has greatly affected the sales of NVIDIA’s graphics cards, and the behavior of many miners selling Pascal-based GPUs has also allowed NVIDIA’s GPUs to accumulate, and the new user base has not yet been determined, resulting in a decline in revenue.Another important factor in the recent sharp decline in NVIDIA’s revenue in the past six months is the uncertainty of data center users’ demand for GPUs.The data center business has been highly anticipated by NVIDIA and is the future direction of NVIDIA.However, data center revenue performance has been poor this year. In the first quarter, data center business revenue decreased by 10% year-on-year, 7% year-on-year, and revenue revenue declined from the previous quarter. Data center in the second quarterThe business fell 14% year-on-year, because large customers have chosen to spend cautiously in the face of the macroeconomic downturn.The sluggish game business is one of the important reasons for the decline in data center revenue. Some analysts believe this is an early warning: the downturn in PC and games seems to be hitting NVIDIA’s data center business; data center business is an important boost to Nvidia’s share price.One of the factors, “the game business has driven revenue, and the data center has driven the downward trend in stock prices.” Artificial intelligence is another application business in the data center.In 2014, Huang Renxun led the company’s transformation into AI technology research and development. The company’s research and development platform CUDA enabled Nvidia to successfully deploy GPU accelerators that reduce energy consumption and improve performance in the HPC data center. The GPU’s parallel computing capability is the basis, while HPC andThe outlook for enterprise computing depends on the extent to which AI is applied.Due to the gradual failure of Moore’s Law, the AI ​​industry needs a method of continuously accelerating performance.NVIDIA’s strategic layout of hardware and software in this field, from reasoning software, integrated optimization containers equipped with NVIDIA GPUs to GPU-accelerated cloud platforms, and accelerators equipped with machine learning reasoning capabilities released two years ago, have enabled Invitro to achieveThe core player in the AI ​​field.However, the marketization of new applications is not so fast. Many AI applications are still in the concept stage. There are very few marketed products, and they cannot immediately form the demand for NVIDIA products. Many new products have a certain premium in the initial stage of market entry.If it is not necessary, many customers will take a wait-and-see attitude.Another expected growth point for NVIDIA is the automotive business, whose products can be used in autonomous driving.However, at present, there is also a market-oriented problem in autonomous driving. The real large-scale market application still needs to wait for a year or two, which also makes NVIDIA’s recent growth less than expected.In terms of the macro environment, the weak global economic development, especially the weak demand in the Chinese market, is also an important factor in the decline of NVIDIA’s revenue in the past six months.According to a report released by the World Bank, the global economy has maintained a growth rate of 3.1% in 2018, but as the growth of developed economies slows down and major emerging markets for exporting commodities and developing economies become more gradual, global growth will graduallySlow, it is expected to slow down to 2.9% by the end of 2020.As the world’s largest consumer market, China’s GDP achieved its expected development target in 2018, with a year-on-year growth rate of 6.6%. However, compared with the growth rate of more than 8% in the past, the downward pressure on the economy is relatively high, and the growth rate continues to decline.Possible.Both the global economy and the Chinese economy are not booming.The big environment is beyond the control of a company. Enterprises can only adjust themselves, better adapt to the needs of the market and users, and increase technology research and development and innovation in order to move forward in adversity.In addition, the author believes that the current situation is also an important external environmental factor for Nvidia listed on the US stock market. Policies and regulations, public sentiment, etc. may have a certain impact on Nvidia’s product sales.So, what did NVIDIA do in the face of such a market environment and the problems caused by it?Anti-slip measures From the second quarter of 2019, the revenue and net profit both fell sharply, but the two increased by 16% and 40% respectively. The year-on-year growth rate declined but the chain maintained a certain increase.Speed, indicating that Nvidia’s revenue is slowly recovering.Nvidia’s share price has risen nearly 6% as both net profit and revenue in the second quarter exceeded Wall Street expectations.NVIDIA’s business revenue is mainly from games, data centers, professional visualization, automotive and edge computing, among which the game business and data center business contribute most of the revenue.In the second quarter, the growth performance of NVIDIA games before the change, breaking through 1.3 billion US dollars, accounting for more than 50% of total revenue, and the growth rate of the chain is also good, becoming a bright spot; data center business realized revenue of 655 million US dollars this quarter,It fell 14% year-on-year and 3% quarter-on-quarter; sales in the automotive sector increased 30% to $209 million.The slow recovery of revenue comes from the anti-slip self-rescue measures taken by Nvidia.In an open letter to shareholders, Huang Renxun announced a number of new measures to prevent further declines in revenue and profits, and to promote the company’s development in a good direction.The first is the adjustment of the game business.From the financial report, the game business rose and returned to the scale of more than 1 billion US dollars. The performance-adjusted revenue and net profit exceeded market expectations.In the gaming business, NVIDIA has introduced new GPU and graphics products, and its latest release of the RTX 2060 at CES has improved performance by 60% over the previous generation.“The RTX 2060, which is priced at only $349, is the first GPU for the mass market and offers upgrade opportunities for millions of gamers,” Huang said.Compared to the $1,089,2080Ti graphics card, the extremely affordable RTX 2060 not only offers gamers a new option, but if the RTX 2080 doesn’t sell, the RTX 2060 can also drive business growth.In addition, NVIDIA has launched more than 40 new GeForce RTX gaming notebooks, more than double the number of last year, to better meet the diverse needs of users.With NVIDIA’s Max-Q platform technology, OEMs can also make their own thin and light notebooks for game enthusiasts.This is followed by the most important upgrade to the workstation GPU for 10 years.At the end of 2018, Nvidia released the most powerful and world’s first “ray tracing” GPU, the Quadro RTX GPU, which was spent 10 years. Huang Renxun expected it to subvert the existing graphics rendering calculations.So, how strong is this GPU?From the show at the NVIDIA GTC 2017 conference, a DGX-Station supercomputer needs to carry four Tesla V100 graphics cards to complete the computing power required for ray tracing, but now a single GPU can be completed, that is,On a supercomputer that requires super power, a GPU can withstand the strength of four good graphics cards.The newly released Quadro RTX includes three products that offer significant improvements in tracking performance, operational efficiency and hardware.The core architecture used by the ray tracing GPU is the new Turing (Turing) technology introduced by NVIDIA, which accelerates real-time ray tracing by 25 times and is 30 times faster than the CPU.In addition, NVIDIA has released the Quadro RTX server, which combines the Quadro RTX GPU with the Quadro Infinity software to ensure both the power and the flexibility of the architecture to meet the creative needs of some professionals.Quadro Infinity allows multiple users to access a single GPU through a virtual workstation, dramatically increasing the density of the data center.Moreover, the Quadro RTX server comes pre-loaded with leading content creation and rendering software, providing a powerful and easy-to-deploy rendering solution for small device installations as well as for the largest data centers with significant cost savings.The third is the new application of data center GPU in the fields of deep learning reasoning, data analysis and machine learning.In 2018, the gross margin of the NVIDIA data center division was between 75% and 80%, but the gross profit margin of different products spanned a lot.According to Deutsche Bank’s estimates, the low-end DGX kit gross margin is about 37%, while the high-end data center GPU gross margin is about 90%, so how to strengthen the new GPU business in the data is extremely critical.NVIDIA introduces computing technology to global corporate customers through partnerships. Amazon, Alibaba, Google, Microsoft, Baidu, Tencent and other giant data centers and Internet companies such as LinkedIn, Expedia, PayPal, Pinterest, Snap, and Twitter all have purchases.Its related enterprise GPU products.In addition, NVIDIA has spent $7 billion to acquire Israeli network hardware company Mellanox to enhance its data center business.In the second quarter of 2019, the revenue of the Nevada data center’s reasoning end increased significantly compared with the same period of the previous year. At present, the reasoning end accounts for more than 10% of the total revenue of the data center.In the autonomous driving business, NVIDIA launched DRIVE AutoPilot, the world’s first commercial autonomous driving system. The automotive suppliers Continental and ZF also launched an automated driving solution based on this system, which will be put into production in 2020.NVIDIA’s DRIVE AGX kit is said to be expected to work with Toyota. If the cooperation is successful, it will not only increase shipments, but also accelerate the popularity of this new autonomous driving system.In conclusion, Nvidia’s anti-slip measures have played a role. Although corporate revenues are still falling, the decline has decreased, the game business has re-emerged over $1 billion, and data center business is slowly recovering.Now NVIDIA is like a student who has taken the first place by chance. If you want to prove that you are a good student, you need to continue to produce good results..