The biggest beneficiary of the Metaverse? Nvidia, the “shovel seller”!

On November 9, NVIDIA announced at the 2021 GPU Technology Conference (GTC 2021) that it would upgrade its product line to a "three-core" strategy of "GPU+CPU+DPU" . At the same time, it positioned its newly released "Omniverse" platform as the "Metaverse for Engineers" .

Compared with social networks, gaming giants, and software giants, NVIDIA is positioned at the bottom of the Metaverse hardware: GPU, AI, and Omniverse. It is worth noting that the Omniverse platform is the culmination of NVIDIA's software and hardware technologies over the past two decades, and is the technical platform foundation for NVIDIA to create the digital virtual space of the Metaverse.

With the help of the "Metaverse", Nvidia's stock price once broke through $323 during the day, setting a record high, and its market value once stood at $780 billion. As of press time on November 10, although Nvidia's stock price has fallen, it still remains above $300, with a total market value of more than $760 billion, ranking seventh in the United States.

Omniverse Platform

Wall Street Journal previously mentioned that Huang Renxun once described "Omniverse" as "a platform that connects the 3D world to the shared virtual world" and is the foundation of future digital twin technology .

Today, the universe is being expanded, and NVIDIA officially launched Omniverse Avatar and Ominiverse Replicator at the GTC conference today. The former can help developers use the NVIDIA Omniverse platform to create interactive virtual characters that can see, speak, have conversations on a wide range of topics, and understand the natural speaking intentions of real people. The latter can help developers create the large amount of data needed to train AI. It is a synthetic data generation engine for training deep neural networks. The objects include general robots and self-driving cars, that is, they are used for NVIDIA's two major platforms, Isaac Sim and DRIVE Sim.

In terms of its application, Omniver locates the Metaverse of engineers and can truly implement the Metaverse in industrial scenarios through NVIDIA's CPU, CUDA, ray tracing and other software and hardware technologies.

According to Jiao Juan's team at Essence Securities , Omniverse has been widely used in multiple industries, truly implementing the collaboration of the virtual world and serving an economy that is larger than the real world. Today, Omniverse has been widely used in six major fields at home and abroad, including media and entertainment, architecture, product design, scientific sports and simulation, autonomous driving, and industrial robots. Such as the media and entertainment field, product design field, etc. For example, the second season of China's official lunar exploration documentary "Flying to the Moon" uses Omniverse for rendering and production; BMW became the first car manufacturer to use Omniverse to design an entire factory.

Focus on hardware, AI, and underlying technologies

According to Jiao Juan's team at Essence Securities, NVIDIA is currently focusing on the three major components of hardware entry, artificial intelligence, and underlying technology.

• Hardware entry: NVIDIA's "GPU Accelerated Computing Platform" integrates hardware and software to provide major companies with a powerful and secure infrastructure blueprint that supports accurate completion of all implementation work from data center development to deployment.

• Artificial Intelligence: Nvidia has taken a leading position in the field of AI chips. According to a cover story in Silicon Valley, in 2019, 97.4% of AI accelerator instances (hardware used to increase processing speed) in the top four cloud providers AWS, Google, Alibaba, and Azure deployed Nvidia GPUs. Karl Freund, an analyst at Cambrian AI Research, said that Nvidia has a "nearly 100% 100%" share of the AI ​​algorithm training market, and nearly 70% of the Top 500 supercomputers use Nvidia GPUs.

• Underlying technology: NVIDIA's Omniverse platform brings together NVIDIA's technologies, algorithms, and standards in AI, HPC, and graphics over the past two decades. It is the technical platform foundation for NVIDIA to create the digital virtual space of the Metaverse.

"GPU+CPU+DPU" "Three-core" strategy

According to Jiao Juan's team at Essence Securities, NVIDIA is now further enhancing its competitiveness by creating a "three-core" strategy of "GPU+CPU+DPU".

Nvidia's powerful GPU, the released CPU Grace, and the latest Bluefield DPU form Nvidia's latest data center chip roadmap. Nvidia's competition in the chip industry has entered the era of combined punches. Through the combination of three chips, Nvidia achieves differentiation and maintains competitiveness.

Risks remain

It should be noted that Wall Street Journal previously mentioned that although the huge imagination space of the Metaverse can excite investors, the gaming business is still one of the company's main sources of revenue, and this business has huge fluctuations .

Looking back at the mining disasters in 2017 and 2018, the graphics card market at all levels collapsed severely, Nvidia's GPU revenue fell sharply, and the company's stock price also experienced a low period of more than a year in 2018 and 2019.

At the same time, competitors are also eyeing the company covetously . To borrow the words of Nvidia founder Jensen Huang, Nvidia, which is "always 30 days away from bankruptcy," is surrounded by "giant companies chasing the same huge opportunity."

In its traditional business, the gaming graphics card market, AMD has made a comeback and launched the RX 6000 series graphics cards whose theoretical performance can compete with the RTX 30 series; Intel, the king of the x86 era, will also enter the graphics card market in the future, trying to get a share of this market with huge potential.

In addition, Apple has also emerged recently , and the graphics performance of its latest CPU can approach the level of 30 series high-end graphics cards while significantly reducing power consumption. Against the background of the strong performance of M1 Max, high-power Nvidia graphics cards seem less attractive in the field of creative design work, especially in video production.