Shaking the AI industry, Yang Yuanqing and Jensen Huang join forces to have a full dialogue!

According to Zhitong Finance APP, on October 16th, Lenovo Group (00992) held its annual technology event, Tech World, in Seattle, USA. The conference brought together top global AI technology giants such as Jensen Huang, Su Zifeng, Mark Zuckerberg, Satya Nadella, Henry Kissinger, and Amon. Lenovo's multiple collaborations with NVIDIA at the event have attracted high attention from the industry, with both parties announcing further expansion of their cooperation and jointly releasing a hybrid AI advantage set as well as claiming the world's "coolest" GB200 liquid-cooled AI server.

The following is a transcript of the press conference:

Yang Yuanqing: Good afternoon! Thank you all for your continuous attention and support for our enterprise AI solutions. To quickly recap, today we shared our vision of the future defined by hybrid artificial intelligence, showcased our innovations in personal AI and enterprise AI fields, and most importantly, reaffirmed the vision of "Smarter AI for All." And all of this would not be possible without a very important partner. Let's welcome NVIDIA CEO Jensen Huang!

Jensen Huang: Thank you, Yuanqing! I am delighted to be here.

Yang Yuanqing: At last year's Tech World, we discussed hybrid artificial intelligence, and since then, both sides have made a lot of progress. The audience must be eager to hear your thoughts. What do you think is the next step for hybrid artificial intelligence? How is the adoption by customers?

Jensen Huang: First of all, I am very pleased to be here again, announcing a series of important new initiatives with our partner Lenovo. Yuanqing and I have known each other for a long time, and we can even say we have been friends since childhood. We have gone through several computing revolutions together, first the PC revolution, then the Internet revolution, and later the mobile cloud revolution. And now, we are reshaping the entire architecture of the computing field on an unprecedented scale.

What we used to call "programming" has now become "machine learning." Programming was done through CPUs, while machine learning is done through GPUs. It is amazing that programming has spawned software, driving the entire vast industry, and now, machine learning is creating artificial intelligence, which will be the biggest industrial revolution we have ever seen. In the past 12 months, we have seen amazing progress in all industries, with every company, every industry, and every country realizing that their digital intelligence and data can be written and transformed into their national, corporate, or industry data intelligence.

Of course, one recent major event is the introduction of the large model Llama 3 mentioned by Mark Zuckerberg just now, which has truly changed the game. With the emergence of Llama 3, every company now has the opportunity to access AI as long as they have the necessary infrastructure and architecture. With AI computers and AI infrastructure, as well as a very critical software stack, we can transform large models into artificial intelligence Artificial intelligence is closely related to large models. We need large language models, very complex large language models and technologies, but (it is important to remember) large language models are an important part of artificial intelligence. The recent content I saw - the complete architecture developed jointly by us as a company and our partners, is to establish the necessary infrastructure, software stack, as well as best practices and blueprints. This way, we can transform large language models into intelligent entities that truly help us accomplish tasks.

Yang Yuanqing: What are your views on intelligent entities AI and physical AI?

Jensen Huang: Regarding intelligent entities AI, when it comes to artificial intelligence, ladies and gentlemen, we hope to produce millions or even billions of AIs like this. At NVIDIA, we call them "Toy Jensens". They will run around and help you with various tasks. Whatever you need, they will meet your requirements. Broadly speaking, artificial intelligence is essentially robots.

In the future, there will be digital robots, which we call intelligent entities. They have the ability to understand your commands, understand the meaning of commands, break them down into specific actions, use tools, retrieve proprietary information or any information they can access, complete tasks, and take action when necessary. So, they can perceive, reason, and take action, and this basic cycle is the core cycle of robotics. Therefore, we will have information robots, which we call intelligent entities.

We will also have physical robots. These physical robots are essentially AIs that understand the physical world. In addition, we will have intelligent entities that understand the information world. These two types of artificial intelligence - intelligent entities AI and physical AI, will become the cornerstone of the global industry. We will have AI "colleagues", these AI "colleagues" may be good at marketing, or in our company, they can engage in chip design, software programming, verification, or building intelligent entities to assist us in supply chain management.

These intelligent entities work in collaboration with all of our employees, greatly enhancing our productivity. Essentially, what we want to achieve is "superhuman" productivity, right? Therefore, all employees will receive timely support from these intelligent entities, thereby increasing productivity. Next, we will also do this. And the biggest opportunity in this is industrial AI, which is closely related to robotics technology, and we are collaborating with Lenovo in this area.

Yang Yuanqing: In terms of intelligent entities AI and physical AI, I also see the same trend. To seize this opportunity, Lenovo and NVIDIA will have a major announcement today, the "Lenovo Hybrid AI Advantage Set".

This is an end-to-end AI platform for developing and deploying AI in the new era. It is based on industry-leading infrastructure, including AI devices, AI servers, storage, as well as edge computing, public cloud, private cloud, where enterprises store, clean, and organize data. Data algorithms and accelerated computing capabilities drive the realization of enterprise AI, optimizing processes, making better decisions, and increasing productivity In all of these elements, services play a crucial role - managing the entire lifecycle of infrastructure including design, deployment, expansion, and maintenance, enhancing data analysis and governance capabilities, providing consultation and tuning for AI models, and expanding AI software ecosystems for customers. Therefore, based on the concept of a hybrid AI factory, we also provide the Lenovo AI Application Library. Our strategy is to combine modularity and customization to quickly respond to customer needs and tailor solutions for them.

All these areas together form the complete system of the "Hybrid AI Advantage Set" that Lenovo collaborates with NVIDIA on. We will work hand in hand with you to make the "Hybrid AI Advantage Set" more mature and complete. So, would you like to talk more about how NVIDIA's technology is helping to build this platform?

Jensen Huang: Yuanqing, when you and I first met, it was during the PC revolution, when the PC architecture was very simple, with a CPU, operating system, and applications. Of course, that computing model was revolutionary at the time, but compared to today, it was too simple. The birth of artificial intelligence technology took a very long time for the entire industry because it is extremely complex. Its entire architecture includes a true supercomputer as the computing infrastructure, running algorithms and distributed computing in a computing structure that includes NVLink/Switch, InfiniBand, or Spectrum-X networks. This distributed computing capability is achieved through very complex software and becomes efficient.

So that's the computer. Above it is the large language model, which may be the new operating system in the world of artificial intelligence. Above the large language model, there are many other application libraries. These application libraries, in a simple way to understand, are like helping a new AI colleague onboard. He is an AI colleague who is there to perform tasks, and your task is to create data to correct and help the AI onboard. He is your AI agent, and you need to teach him very specific skills.

You provide them with information so they can do their work. You evaluate them like you evaluate employees. You deploy them, set up barriers for them, protect them, ensure that the functions they perform are very much in line with what they have been taught to perform. And Blackwell makes the intelligent agents better over time. This whole set of application libraries is fully accelerated and runs on GPUs.

This makes it possible for every company to introduce large language models and turn them into their own intelligent agents to perform specific tasks you want them to perform, such as customer service, AI database retrieval, and many different types of applications.

Yang Yuanqing: Indeed, NVIDIA does provide a lot of value in this regard. There are corresponding platforms at various levels. (We need to) meet higher goals for customers - they not only expect more powerful accelerated computing capabilities but also aim to achieve energy efficiency goals. So, how does NVIDIA help meet such customer needs?

Jensen Huang: In the early stages of the new wave of computer revolution, your best choice is to accelerate your roadmap. The reasons are as follows: when performance can double or triple - and this is annual growth - we can effectively reduce the cost of AI, reduce the workload and energy consumption of AI, while also increasing the revenue-generating capabilities of AI These new computers are like efficient operating factories.

Unlike data centers that store files, these data centers are AI factories that produce tokens. You would want the tokens, which are actually artificial intelligence, to be produced at the highest possible speed.

If we can double or triple the speed every year, we can help you reduce costs and increase revenue. So, this is very important. We have the ability to do this because we have built the entire AI factory end-to-end, from CPU, GPU, NVLink/Switch, InfiniBand Switch, network chips to Ethernet switches and networking technologies.

We have the ability to build the entire data center and AI factory end-to-end, and we also have the ability to develop software end-to-end, from start to finish. That's why we can build new AI factories every year, doubling performance, reducing costs, and driving the development of AI at the fastest pace and democratizing it.

Yang Yuanqing: All of our R&D teams are dedicated to the mission of sustainable design.

Jensen Huang: Now, speed is sustainability, speed is performance, speed is energy efficiency.

Yang Yuanqing: Undoubtedly, NVIDIA has done an excellent job in sustainability. Lenovo also brings core technology, and we have been in a leading position in liquid cooling technology for ten years, now in the sixth generation, which is our Neptune liquid cooling system technology. So, let us now demonstrate a product using the sixth generation Neptune liquid cooling system.

Jensen Huang: Of course, please. You know, Yuanqing, the efforts Lenovo has put into building high-performance computers over the years have paid off. Isn't that right?

Yang Yuanqing: Absolutely. We have been researching water cooling technology for over a decade. This is the ThinkSystem SC777 system designed by Lenovo, equipped with NVIDIA Grace Blackwell GB200. It features a 100% liquid cooling design, so it does not require any fans or dedicated data center air conditioning. The system can be installed in a standard rack, using standard power, so customers can purchase one tray at a time. It also includes NVIDIA NVLink interconnect technology and supports NVIDIA networking and NVIDIA AI enterprise software. What are your thoughts on this?

Jensen Huang: Yes, this is really beautiful. For an engineer, this is very sexy.

Yang Yuanqing: Thank you, Jensen, for our important partnership. In addition to our collaboration in enterprise data centers, we have also partnered with NVIDIA in the field of automotive computing, launching the NVIDIA DRIVE AGX Thor platform with the new Blackwell architecture. Now, let's showcase Lenovo's smarter and more powerful AI computing DCU together with Jensen Jensen Huang: Yuanqing, do you know how amazing this is? The architecture of this computer and that computer is exactly the same. That computer is used to create the "brain" running on this computer. Because cars are the most produced robots in the world.

Yuanqing Yang: Yes, that's right, robots on wheels.

Jensen Huang: It can perceive the world, think of action plans, and control the steering wheel like a robot. Therefore, this will be the first batch of robots with the highest production volume in the world. But in the future, we will have all kinds of robot systems, and this special computer can be used for all these systems. So, yes, very exciting. Thank you.

Yuanqing Yang: Anything else to add?

Jensen Huang: Go Lenovo! Thank you.

Yuanqing Yang: We are delighted to cooperate with NVIDIA. Thank you, Jensen