News classification
Contact us
- Add: No. 9, North Fourth Ring Road, Haidian District, Beijing. It mainly includes face recognition, living detection, ID card recognition, bank card recognition, business card recognition, license plate recognition, OCR recognition, and intelligent recognition technology.
- Tel: 13146317170 廖经理
- Fax:
- Email: 398017534@qq.com
In an AI keynote speech at Computex's Computex international computer show in Taipei last week, huang renxun, the chief executive of the "godfather of AI", officially unveiled the design of the Max Q laptop.
The Max - Q design from four aspects: the best energy consumption than the GPU Settings, all-round optimization and efficient cooling solutions and quality setting a new generation of high efficiency power supply design. According to the nvidia official introduction, Max - Q design game notebook can not only accomplish originally a third of the thickness, 1/5 at the same time also can reduce weight, performance increased by 3.3 times.
After lei feng network reported that as of April 30, according to results of the first quarter of fiscal 2018 first fiscal JiYing weida data center business revenue of $409 million, more than double compared to the same. The business of data centres almost tripled last year, Mr Huang said. The auto maker's first-quarter revenue was $140 million, up about 24 percent from a year earlier. In addition, the game business first fiscal revenue was $1.03 billion, year-on-year growth of nearly 50%, accounting for 44% of the company's overall revenue, is still the nvidia's main source of revenue.
In may, nvidia released five revolutionary products at the GTC conference: Project Holodeck, Volta, nvidia GPU cloud, Xavier DLA, and the Isaac robot simulator. VR, cloud computing, AI, autonomous driving, robo-training and other popular areas have fallen.
Nvidia has positioned itself as an AI company since 2016, and the data center business has maintained a high level of growth. But as the biggest source of revenue, nvidia is clearly not ready to give up the game market, and Max Q's release is the best proof.
So how does nvidia plan for its future?
After the event, huang received numerous media interviews. Huang renxun gave his answers to the outlook for the future data center business and gaming business, the assessment of TPU, and the view of the Chinese AI industry.
The following is a live interview in shorthand, and the lei feng network (public no. : lei feng network) did not change the meaning of the compilation.
Q: will the data center replace the game as the highest revenue service in nvidia?
Huang renxun: I don't want it. I believe it will be hard, because I believe the time has come for the GPU. This is not about whether nvidia's products are popular, it's about logic and physics. The reason I don't want this is because I love gamers, and you know that the future of the game market is not going to shrink, and that logic is clear. Now there are 200 million GeForce game users in nvidia, and I believe everyone will play games in 20 years.
No one knows how to play games in the future because everyone has their own "virtual reality". Sometimes you will go to your own parallel universe, and I will be in my parallel universe, and these different worlds are interconnected. We can study and work in it. The virtual reality, the video game, these are the areas of it. So, I believe that within 20 years, everyone will be a game player for the video, or the inhabitants of the parallel universe of VR.
, of course, we are now the fastest-growing data center operations, and we think it would be a very big city, because the AI will often walk into every person's life.
Q: apple, Google, and even tesla are developing their own AI chips. What do you think the market will be like in the future? Will there be competition for nvidia?
Huang renxun: the first question, I think AI is clearly a very big market, and in fact it will affect every influential industry. It will affect every company, affect every country, and affect the way we build computers. Now, every computer shipments billions of units, including mobile phones, PC, server, data center and cloud, and there are billions of cars and IOT devices in the future.
Everything involves AI, right? So we're talking about computers, mobile phones, cars, the Internet of things, all of which have AI. At any time, the more companies start to join the AI camp, the faster the AI will come. For five years, when we started to invest in AI, the AI didn't get as much attention as it does today. But the more people start to do it, the faster we grow. Because the more people go to AI, the more people talk about AI, the more people use AI, and we'll grow better.
The AI industry was close to zero, and AI will be the world's largest industry. No matter where you are on this track, the sooner the better. I believe that a lot of companies are working on AI because it's very important for them. Our strategy and contribution must be unique. Our strategy is really quite simple. We have a large team of engineers focused on AI computing, actually the largest in the world.
We are very good at building computing platforms. Very competitive in the field of software, and we are very good at it, our strategy is to create a platform, then grab the other companies in the world before entering, and let every company in the world, every industry can be used. Our goal is to democratize the AI. We're not going to build for ourselves. We want everyone to use the platform. Whether others prefer to use our platform or build their own platform, it's ok for me. There are so many companies in the world, and I believe we will continue to be the best AI company in the world.
Q: what do you think the company should be responsible for when developing a technology?
Huang renxun: our products are mostly tool-like products, like PCS and bicycles. In fact, our company is a tool, but unlike most tools, it allows us to see and experience the future, and we do that by accelerating. If we make an event 100 times faster, then researchers, doctors and artists can see the future. And that's what the time machine was, and we designed this tool to make people more imaginative. Some people are trying to build engine equipment to help us cope with cancer, and others help us to understand society and communities. So I think we are creating technologies that are good for all of humanity.
Q: how do you see the competition between gpus and TPU?
Mr. Wong: three letters doesn't mean anything. GPU applies not only to the graphics (calculated), can also be used for the dynamic simulation in the process of developing new drugs, quantum mechanics, quantum chemistry, by using difference equation to simulate the influence of the car crash, computer vision, and improve the image quality of filmmaking. The architecture of the GPU makes it very good at deep learning.
So my job is to make sure that nvidia's GPU is not the same GPU that other manufacturers advertise. For example, you can't directly learn from the gpus on the integrated graphics card, nor do you directly learn from the gpus in your smartphone, while nvidia's gpus can. Our gpus are designed to handle general-purpose computing and high-throughput linear computing, which is what TPU does. I think that TPU and GPU are essentially the same things, both processors, and have their own strengths and limitations. Our gpus are stronger and more flexible because we do a lot of things, so all the AI networks you can think of at the moment can be perfectly run on our gpus. TPU is limited because it has limited precision and is designed to be simpler, which was decided at the beginning of the design.
Again, both are chips.
Q: Japan's softbank recently acquired some of nvidia's stock, and it values the possibility of nvidia's development in the AI sector. So does this account for the future business plan of nvidia? What is the future of nvidia? What are your expectations?
Huang: a lot of people don't know this, but masayoshi son, softbank CEO) is the only one I know is, predicted every time the arrival of the computer revolution. You know, he brought Microsoft into Japan, brought yahoo into Japan, brought apple into Japan. The PC, the Internet, the mobile Internet, every time he's in the front line of the revolution. Today, the AI revolution will be the biggest revolution. Son and I have known each other for a long time, we are friends, I think he is investing in nvidia is to see the future of our company, but also for his own future layout.
Q: some people say that Russia and north Korea are trying to build AI in their own right, even trying to poach key technologies from the west. What do you think about this?
Mr. Huang: no matter what decisions we make, we won't stop them from investing in AI. So we have to accelerate the investment in AI, right? That's why we have to democratize AI. I believe there are more people here than there are men, and that's why we're here. , of course, this is my expectations of rising, and growing presence in the life science and technology, and constantly improve, because most people are good people, now more secure than in the past.
It is safer now than it was a thousand years ago, and I believe that humanity is generally well-meaning. So we can get all these good people to get the best technology. I believe this is often done in a good way, but we still face obstacles. In both nuclear and biotechnology, the development process has been accompanied by the sound of protests against their harms. We know that technology is quite open, and one of the reasons they say that is Einstein. In Einstein's letters, he all comments are widely attention to science and technology, he expounded his concerns and could face a crisis of science and technology, he put forward the idea of powerful technology, the future one day, the power of science and technology can be used by the world, even those who intended to use science and technology caused the destruction of the destroyer.
I think this relatively mild approach to the development of today's society and the public debate on technology can help ensure that technology is properly utilized.
Q: what do you think of China's AI? Can you give me some advice?
Huang renxun: China's AI is developing, in the process, full of imagination. The number of Chinese papers, start-ups and researchers focused on AI has been growing rapidly. I think China's interest in AI is not even on a level with the rest of the world. Because the Internet industry is the most important industry in China. Some of China's Internet industry leading enterprises, such as alibaba, tencent, baidu, etc., and now they have a good understanding on deep learning, understanding, which is the basis for AI is also in place.
I think China has a very unique environment, very interesting culture, and it has the ability to move fast. In China, access to capital is easy, so investment levels are high. I think China's environment is very unique, every country must realize that China is a single, powerful technology, jungle that is never seen in this world.
Every country must invest and improve its capabilities. The development of AI, the time is not me. The ability to develop AI technologies will benefit society and the country. And these companies have never stopped exploring the AI.
As you know, there are two types of companies in the AI field: the company that develops AI and the company that USES and USES AI. If you are a wealth company, you want to develop AI and develop your own product with AI. If it's a manufacturing company, you need to develop AI and use AI. Most companies in the world today are in a hurry to use AI. That's why we work to help them understand the AI. Nvidia has opened a school of deep learning for these companies, which is expected to attract nearly 100, 000 students this year. In addition, we are working with all of SAP's corporate giants, so we can bring the enterprise capabilities of AI to everyone.