The market value is four times that of Intel, so why is Nvidia?

The biggest beneficiary of the data dividend

Benefiting from the outbreak of the cloud computing ecosystem, in the third quarter, Nvidia’s data center revenue was US$2.9 billion, an increase of 55% from US$1.9 billion in the same period last year, becoming the strongest driver of revenue growth.

If it really enters the meta-universe era, data will exponentially explode, and the demand for high-performance chips will be greater by then, and NVIDIA has already taken the lead in layout.

From 750 billion U.S. dollars to 820 billion U.S. dollars, in just three days, Nvidia’s market value rose by 70 billion U.S. dollars. All this happened after the third quarter earnings report was issued.

On November 18, Beijing time, Nvidia announced its third-quarter financial report as of October 31, 2021. The financial report showed that this quarter’s revenue was US$7.1 billion, a year-on-year increase of 50% and an increase of 9% from the previous quarter, especially for games. , Data center revenue hit a record high. Affected by the performance, Nvidia’s share price rose 5% after the market.

Affected by the short supply of high-performance chips caused by the data center and digital currency boom in the past two years, plus the replenishment of Meta, Microsoft and other companies in the meta-universe field this year , NVIDIA has stood on top of the “data explosion”. As a “seller” at the core of data processing, Nvidia’s stock price has risen from $30 in 2019 to over $300 this year, an increase of 10 times, and its market value is already 4 times that of Intel.

The market value is four times that of Intel, so why is Nvidia?

(Nvidia’s stock price trend chart for the past 5 years)

The company that once started with computer graphics chips has a market value of US$820 billion in just three years. The growth rate of Nvidia’s market value can be described as a rocket increase, but compared with established technology giants such as Apple, Google , and Microsoft, What’s the secret that this company has been able to stand on the air frequently? Faced with the continuous investment and catch-up of Intel and AMD in the GPU field, how will NVIDIA respond?

Data center and game business both surge

In Nvidia’s third-quarter financial report, its $7.1 billion revenue was mainly composed of four major sectors: games, data centers, professional visualization, and automobiles . The core reason that investors are optimistic about Nvidia lies in Nvidia’s breakthroughs in gaming and data center business . These two businesses have not only maintained a rapid growth momentum, but are also the main drivers of revenue. Both are higher than market expectations and have become “two axes” to promote the increase in NVIDIA’s market value.

The market value is four times that of Intel, so why is Nvidia?

Gaming is the core of Nvidia’s business. According to data, Nvidia occupies 83% of the current global game graphics market share. In this field, it has obvious product competitive advantages compared to rivals such as Intel and AMD, which contributed to its third quarter game revenue. The performance of US$2 billion in the same period last year increased by 42% compared to US$2.27 billion in the same period last year and a year-on-year increase of 5% compared to the previous quarter.

However, Nvidia’s financial report did not give a detailed description of the specific growth of its game business customer group. It is understood that this includes not only gamers, but also a large number of virtual digital currency “miners”. Since Bitcoin, Ethereum and many other types of digital virtual currency “mining” require higher computing power graphics card support, leading to NVIDIA’s Gaming graphics cards have been in short supply in the sales market for the past year.

However, in the financial report, Nvidia collectively referred to the sales growth of graphics cards as the “game” business, but the actual proportion of the increase due to “mining” users is unknown.

From another point of view, it seems to explain this problem. Although Nvidia saw the advantages of its graphics cards in the virtual digital currency field, it also launched a graphics card product CMP dedicated to mining in March this year, but its sales are very average. In the first fiscal quarter of this year, Nvidia CMP’s sales were 155 million U.S. dollars and reached 266 million U.S. dollars in the second fiscal quarter. However, CMP sales in the third fiscal quarter were only $105 million, a decline of 60% from the previous quarter. Nvidia also expects that the product’s sales in the fourth fiscal quarter will drop to a “insignificant” level.

The reason for this is, on the one hand, the high cost performance of game graphics cards, and on the other hand, it is also caused by its high circulation rate in the second-hand market. Nowadays, the supply and demand imbalance caused by the shortage of global chips, although most gamers have very strong demand for game graphics cards, there are also many “miners” who buy NVIDIA game graphics cards, which has also led to NVIDIA’s continued profitability in the field of game graphics cards.

Next is the data center. Nvidia’s data center revenue in the third quarter was US$2.9 billion, an increase of 55% compared to the US$1.9 billion in the same period last year, becoming the strongest driver of revenue growth. In June of this year, NVIDIA received another order from Microsoft. Its Azure cloud platform uses virtual machines supported by NVIDIA A100 GPUs to become the focus. Due to the increasing trend of remote office under the influence of the epidemic, it is good for cloud service providers, with higher computing power and Stronger AI processing capabilities are inseparable from the advanced technology in the GPU field, which is the reason why Nvidia’s data center business continues to grow.

Colette Kress, Chief Financial Officer of Nvidia , mentioned that the growth in this area is mainly driven by the sales of “super customers” who demand GPUs. Among them are customers such as Amazon AWS, Microsoft Azure, Google Cloud, etc., due to the rich GPU product line of Nvidia , The supporting programming environment is more mature, so it is widely praised by AI developers. In recent years, due to the rapid growth of artificial intelligence tasks such as voice and graphics data processing, it has also promoted the rapid growth of NVIDIA’s data center sector, which is expected to surpass games to become the largest business.

In addition, the professional visualization business has also grown rapidly, with revenue in this area of ​​US$577 million, a year-on-year increase of 144% over the previous year and a year-on-year increase of 11% over the previous quarter. But in the automotive business, revenue in the third quarter was US$135 million, a year-on-year increase of 8% compared with last year, and a year-on-year decrease of 11% compared with the previous quarter. Overall, the professional visualization business and the automotive business are currently on the rise, but they have not yet Occupying an important proportion of revenue, NVIDIA tried to walk with “four legs”, but these two legs have not yet grown up.

However, what is worth paying attention to in the financial report is the increase in its operating expenses and net income. Compared with the same period last year, operating expenses increased by 25% to 1.96 billion US dollars, while net profit increased by 84% compared to the same period last year to 2.464 billion. The dollar, which also reflects Nvidia’s strong ability to make money.

There were Intel and AMD before, and Amazon and Ali after that

The rapid growth of Nvidia’s share in the field of graphics chips has led to the current “dominant” market structure in this field. In the face of such competition, other competitors including Intel and AMD have begun to challenge Nvidia.

In 2019, Intel acquired Habana Labs, an Israeli semiconductor company, for US$2 billion to strengthen Intel’s technical strength in data centers, especially in artificial intelligence products. In March of this year, Intel demonstrated its first exascale computing GPU product-Ponte Vecchio, which has more than 100 billion transistors and can perform four trillion floating-point operations per second. It is made up of 47 chips called “Velcro”, including logic, memory and I/O controllers, and is manufactured using processes of different nodes. It is a master of Intel’s advanced technology at this stage.

According to Intel’s official statement, this product is different from any of its past products, “almost all parts have been redesigned and reorganized”, and “it took only two years to be born after the concept was put forward, but under normal circumstances it requires A few years’ time”. Its main competitor is Nvidia. Officially, the Xe HPC core in Ponte Vecchio GPU has 8 vector engines (512-bit) and 8 matrix engines (4096-bit), which can play a role in the field of high-performance graphics computing.

The market value is four times that of Intel, so why is Nvidia?

(Chart from Intel)

Obviously, Intel has high hopes for the GPU field. According to the official description, with the emergence of Ponte Vecchio in 2021, it is expected to start catching up with its competitors in the HPC GPU field (obviously NVIDIA) this year. The product will be used as a supercomputer accelerator, but the specific effect is not yet known, but it seems that Intel has also begun a counterattack in the GPU field after years of silence.

AMD is constantly challenging Nvidia’s dominance in the consumer gaming graphics card market. Although from Fury X in 2015, RX Vega in 2017 to Radeon VII in 2019, AMD’s challenges almost all ended in failure. In November last year, AMD launched another offensive and launched the RX 6000 series of graphics cards, trying to shake Nvidia’s dominance through price, but as of now, Nvidia still firmly controls 83% of the market share, while AMD occupies 17%. Market share, in the consumer-level market, due to Nvidia’s leading advantages in R&D investment and technology, it is very difficult for AMD to catch up or even surpass.

The reasons for this are very diverse. Taken together, on the one hand, due to Nvidia’s first-mover advantage and continued heavy investment in technology research and development in the GPU field, it is difficult for competitors to find opportunities to surpass its product advantages. It can also be seen from the research and development of the financial report that in the past ten years, Nvidia’s expenditure on research and development has exceeded that of its main rival AMD, and this gap has continued to widen. According to financial reports, in the first three quarters of this year, Nvidia’s R&D investment reached US$3.798 billion, while AMD’s R&D investment was US$1.883 billion. The former is twice that of the latter.

The market value is four times that of Intel, so why is Nvidia?

(Data comes from their respective official websites)

On the other hand, the scale effect and ecological effect of the chip industry are also very important. A better programming environment and long-term developer cooperation have created NVIDIA’s professional barriers in the graphics card field. Although the gap between Amazon and Alibaba and other emerging challengers in the hardware field will gradually narrow, how can they build something similar to NVIDIA? Ecological barriers are a major challenge that needs to be faced.

Standing at the starting point of the “virtual world”

Nvidia launched the world’s first GPU-GeForce 256 in 1999. Since then, the company has been running wildly in the GPU field and successfully transformed in 2008, creating a CUDA programming model and Tesla data for service centers. The central GPU has since been no longer limited to the PC market.

Due to the GPU platform, it is better at large-scale parallel data processing and rendering graphics than the CPU. This has laid an important foundation for NVIDIA’s multi-point penetration in the fields of games, AI and robotics, data centers, and digital twins. Taking advantage of the combination of software and hardware, Nvidia has successfully entered the AI ​​service provider from gaming hardware service providers and will soon enter the meta-universe. Nigel Toon, co-founder and CEO of Graphcore, a British artificial intelligence chip design company, once commented: ” Nvidia’s GPUs are effective because they create software libraries, frameworks and optimizations that are better than their opponents.”

This advantage has become Nvidia’s entry ticket to the AI ​​field and “Meta Universe”. According to the latest official data, the number of its developers is close to 3 million. CUDA has been downloaded 30 million times in the past 15 years, and the number of downloads in the past year Reached 7 million, and almost all AI computing-related technology companies will use Nvidia’s GPU as their hardware platform. Huang Renxun said that not only general developers, but more than 25,000 corporate customers including Microsoft, Samsung, Snap, etc., are using NVIDIA’s AI reasoning platform, which proves NVIDIA’s market appeal and ecological barriers on the software platform.

In addition, Nvidia will target the “meta universe”, the hottest field at the moment. Nvidia tries to create real scenes, real sounds, and a more convenient tool platform for making these, so it has the Omniverse Avatar platform, NVIDIA Quantum-2, Omniverse Replicator and other products that have been publicized not long ago.

The market value is four times that of Intel, so why is Nvidia?

(Picture from Nvidia)

Omniverse Avatar is part of NVIDIA Omniverse, a virtual world simulation and collaboration platform for 3D workflows. This is an interactive character with ray-traced 3D image effects. Everyone can create their own virtual avatar, they can see , Speaking, talking about various topics, and understanding and expressing intentions reasonably. As part of NVIDIA’s creation of the “meta universe”, its appearance is to further bring collaborative robots and virtual assistants into the virtual world we are accustomed to today. The voice AI, computer vision, natural language understanding, recommendation engine, facial animation, and image elements used by Omniverse Avatar are all coordinated by software and hardware independently developed by NVIDIA.

In addition, Nvidia NVIDIA Quantum-2 is a hardware booster to push Nvidia forward. According to Huang Renxun, the founder and CEO of NVIDIA, “The ultra-high performance, wide accessibility, and high security provided by Quantum-2 are exactly what cloud computing providers and supercomputer centers need.” It used to be costly. A few months of simulation can now be completed in just a few days. When NVIDIA creates its own meta-universe software, the advantages of the hardware part are further strengthened.

Even the emergence of Omniverse Replicator is to attract more developers to join. Through this synthetic data generation engine for training deep neural networks, they can help developers create the massive data needed to train AI more quickly.

Compared with established rivals such as Meta, Microsoft, and Apple, Nvidia, which stands in front of the “Meta Universe”, has already taken the lead. Because for Nvidia, the more data, the higher the profit.

Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2021-11-22 12:35
Next 2021-11-22 12:43

Related articles