Where is AI going?We saw these trends in GTC 2022

There are not many conferences that can be called the "Spring Festival Gala", and GTC is considered to be one of them. Not only are he attracted a large number of game enthusiasts, but also more viewers come to NVIDIA's "AI new dishes" and "hard dishes" on NVIDIA each year.

The application of GPU has brought new variables to the AI ​​industry, and AI has also become a new strategic export of many industries. Faced with the trend of AI industrialization and industrial AIization in full swing, few people are not curious about what the latest trends in the AI ​​world, and the GTC conference is often the opportunity to observe the latest AI technology, software and hardware, and industry applications. Whether it is the latest "performance monster" for AI violence, or interesting applications such as Huang Renxun digital people, or solutions for AI landing scenes such as new drug research and development, scientific calculations, Yuan universe, and digital twins, "including AI quantity "A high GTC conference can always bring a little inspiration to everyone, or refresh the awareness of the AI ​​technology and AI industry.

Then, the audience, we started from the GTC Spring Festival Gala to see what dishes the NVIDIA and even the AI ​​industry chain in 2022 are all thousands of industries and the masses.

AI stage built by NVIDIA

More than one event known as the "Technology Spring Festival Gala" is equally noticeable. Therefore, it is necessary to explain why we want to use the GTC conference to use the AI ​​industry and industry AI of the pulse 2022.

The first is technology. GPU is known as the "Pearl in the Crown" in the field of chip design, which has played a decisive supporting role in AI computing power. It is also the core product of NVIDIA. Therefore, GTC can get more information about new products every year.

The second is the application. To sell GPUs in NVIDIA, it is natural to create a practical application scenario for potential users and customers. Therefore, GTC also has a large number of AI applications every year, bringing a lot of inspiration to developers.

The third is industry. The GPU is the foundation of the AI ​​industry chain. Every year, GTC often has a number of middle -level industrial chain, such as automobile manufacturers, server manufacturers, cloud service providers, etc. to start docking to launch more targeted products and solutions. Have more important reference value.

So, what are the changes in the GTC 2022 stage to show us?

GTC 2022's AI dance

I have to say that the lineup of the GTC 2022 is still very heavy. There is a feeling of gathering "Shenlong" to summon the future. Other activities and workshops are also wonderful. From the perspective of macro trends, we will interpret what new technologies, new trends, and new situations do GTC 2022 show.

First, AI applications are fully blossoming. If AI still has participated in some key areas such as the Internet, autonomous driving, and film and television production before, then many issues of GTC 2022 explained that from developed countries to the third world, from scientific research to real economy, When you go to technology companies, no one does not care about AI, and innovation around AI has blossomed worldwide. For example, one of the important issues of the conference is how emerging economies in Africa, India, Israel and the world use AI to improve the efficiency of supply chain; The AI ​​innovation exchanges have become extraordinarily colorful. The industrial element universe, the digital twin of the earth, the world's first AI ranch Betsy, etc. AI has changed from the case -level "possibility" to all walks of life and countries in all walks of life, countries and locations from all over the country. "Must."

Secondly, basic innovation slows down, but its industrialization capacity is enhanced. Time comes to 2022, the highlight of the AI ​​field is still a big model. However, the algorithm innovation, product innovation, and tool upgrades around large models are still continuously developing. In the GTC conference, a large number of activities around large models can be seen. For example, how to allow all developers to access hundreds of billions of parameters Customized language model LLM, NVIDIA supports large models of large models in biomolecules and protein. In general, this also means that more people have begun to use new technologies to participate in industrial intelligence and transform technological value into commercial value. Compared to the burst AI innovation in the laboratory, more and more AI innovation that integrates fireworks will be integrated into the fireworks.

Finally, ecological partners are fused and digital infrastructure upgrades. On GTC 2022, we can see a large number of familiar companies, manufacturers and AI hardware solutions. Calculating the positive hugs of AI in the calculation of the industry and digital infrastructure means that AI accelerates the feasibility and broad prospects of the integration of the industry into the industry. On the one hand, NVIDIA has launched a large number of GPU clusters for large machine learning infrastructure to adapt to the large -scale parallel AI training and reasoning tasks. At the same time, green low -carbon has become a key capabilities for data centers. At the conference, it discussed the high -density rack -grade -grade -grade liquid cold server based on NVIDIA liquid -cold GPU, providing green and low -carbon computing capabilities for the digital economy.

In general, the various signs of the GTC 2022 show that AI has become a highly available basic technology and infrastructure, and the breadth and depth of integration with the industry have reached unprecedented levels. It can be expected that AI industrialization and industrial AI will be accelerated by the promotion of various roles of the industrial chain in just a few years. AI attack in the industry vertical and horizontal

As mentioned earlier, after the GTC 2022, the intelligent change brewed in the industry scene will definitely enter a new chapter. On the base of AI hardware, the relevant industrial chain will also change quietly, penetrate the industry scene, and complete the value conversion of AI hardware technology innovation in specific products.

In China, the movement of NVIDIA cooperative companies is a vane of the underlying hardware to drive industrial change. The server products certified by NVIDIA often have good performance.

Taking Ning Chang as an example, as a start -up server manufacturer, closely coordinated the global AI industry chain, and completed the "one horizontal and one vertical" industry in a short period of time:

"One horizontal" refers to the breadth of AI products. At present, Ning Chang has been recognized by head users in the Internet, telecommunications, finance and other industries, and the coverage of scenarios and market shares have continued to expand. Calculate Market Analysis Report "). Thanks to the hard power of the Ningchang AI server's own products, it has created a number of "NVIDIA-CERTIFIED Servers" (NVIDIA certification server) artificial intelligence servers, and it has performed well in the selection test of industry users. On the other hand, it participated in the NVIDIA certification system ecosystem and received enterprise -level support from NVIDIA software stack. As MANUVIR DAS, the person in charge of the Nvidia enterprise calculation, said that through the NVIDIA verified server, it has the ability to empower the world -class AI and further open up the market -oriented channel of Ningchang.

"One vertical" refers to the vertical market with the most chance of cultivating, consolidating its core product power, and forming a benign closed loop from research to use. In response to the "carbon reduction" demand of the computing center, Ning Chang launched a liquid cooling product covering the GPU server, taking into account the computing performance and green low -carbon; in the intelligent industrial intelligence, relying on advanced artificial intelligence servers can not only meet the conventional AI reasoning 2. Online recommendation system needs, and can also build GPU high -speed clusters to require higher workloads, such as dialogue artificial intelligence. Through the deep cultivation and satisfaction of multiple intelligent scenarios, Ning Chang can successfully complete the "product research and development-market recognition-commercial monetization" virtuous circle.

This "one horizontal and one vertical" continues to stretch, representing the value space of AI industrialization and industrial AIization.

It can be seen that AI has completed "self -evidence". Today, when it comes to AI, few people question its value, but hopes to seize this is the most critical growth variable. Entering the new stage of development, AI needs the "other cards" that have landed in the industry and the scenes, which are what the GTC conference and their ecological partners are doing.

From the "Spring Festival Gala" in the scientific and technological world, to the light of the smart China, the stage of the AI ​​era is being lit by Ning Chang.

- END -