Concerns about semiconductors are growing.

The storm of semiconductor chip bargaining and price reduction has expanded, and the quotation of MCU (microcontroller) has also fallen sharply, especially in consumer applications. Taiwan Semiconductor Manufacturing Co Ltd's three major customers reduced the order volume of the news, let the market is also very frightened.

NVIDIA Corp (NVIDIA), as the leading company in the global semiconductor sector, has not been immune from the stock price performance in this round of slump in technology stocks.

Although the latest quarterly results released on May 26th showed strong business growth, the latest closing price of $145.23 was still about 17% lower than the price at the time of the announcement. Its share price was $346.47 at the end of November.

NVIDIA Corp, founded by Huang Renxun in 1993, is a wafer-free IC semiconductor company specializing in the design of smart core chipsets and a leader in graphics processing technology.

Focused on creating products that enhance the human-computer interaction experience of personal and professional computing platforms, it is still one of the largest video card chip manufacturers in the world.

NVIDIA Corp's graphics and communication processors have been used on a variety of computing platforms, including personal digital media PC, commercial PC, professional workstations, digital content creation systems, notebook computers, military navigation systems and video game consoles.

When asked what is the most important ability of an individual?

Huang RenxunAffirmed: "it is my vision, the way I see the world." "

NVIDIA Corp was founded as an accelerated graphics processor. At the beginning, the target market was games.

But with the continuous development of society, Huang Renxun realized that a single-function image processor will not become a sustainable business in the future, and the company must make changes.

In its latest quarterly results, NVIDIA Corp's revenue reached a record $8.29 billion, up 46 per cent from a year earlier, while net profit fell 15 per cent to $1.62 billion.

Of this total, data center business revenue was US $3.75 billion, up 83% from the same period last year, gaming business revenue was US $3.62 billion, up 31% from the same period last year, professional visualization business revenue was US $622 million, up 67% from the same period last year, and revenue from automobile and robot business was US $138 million, down 10% from the same period last year.

NVIDIA Corp has changed from a niche market to a chip manufacturing giant supported by four major product lines, and its GPU, DPU, CPU, software and systems are all in the market leading position.

Compared with the troubled technology industry, bright financial results are not the most important, the key point isNVIDIA CorpThe strong growth of data center business also reflects the market.The new trend of：AIIt is widely used in the new scene of autopilot and digital twinning.

As for the record performance of the data center and gaming business, Huang Renxun said with great ambition:

"We are preparing for the largest wave of new products in history, with new GPU, CPU, DPU and robot processors coming out in the second half of the year.

Our new chips and systems will greatly promote the rapid development of AI, graphics, Omniverse, self-driving cars and robotics, as well as the many industries affected by these technologies. "

In the investor exchange after the disclosure of the quarterly report, Huang Renxun reviewed the development process of NVIDIA Corp along the way, detailing the challenges and initial intentions of nearly three decades.

When asked about what he wants to do in the future, Huang Renxun said that what we want to do is to find the most difficult but meaningful problems in the world within our circle of abilities, and then solve them.

This is an ambitious manifesto by Huang Renxun about betting on the future, which is more surging than a dialogue around the meta-universe last November.

Smart investors sort out some of the best ideas from this investor exchange and share them with you:

If you want to be a market leader, you must take the initiative to nibble away at your own products

1. When you are in the high-tech industry, when technology is developing so fast, if you don't reinvent yourself, you are just dying slowly.

We realized that we had to make GPU programmable, or we would only be eroded by other competitors. If you want to be a market leader, you must take the initiative to nibble away at your products.

This is the biggest bet in NVIDIA Corp's history.

two。 When the trend of AI breaks out, I think the most important thing is that my understanding of software has changed.

In the past, software programs were written by people, but now software has become written by machines.

The machine can extract from the data, identify the relationship between patterns and data in some way, and learn some prediction models in some way, the whole process is completely automated.

So the computer needs to redefine the software system to adapt to the process of AI.

When this new trend breaks out, it will change all industries.

Therefore, our core is to provide new solutions to these problems.

3. Since 2013, almost all technology companies have seen the trend of AI, but why do we have an advantage over chip companies such as CPU and other competitors?

Because to do accelerated computing, it must be a complete company, from hardware to software, to algorithms, which is very different from CPU's way of thinking.

In the business of accelerated computing, you need a lot of good algorithm engineers.

At the same time, you have to systematically consider the entire system in different ways, including software, network connections, memory, read speed, the coordination between CPU and GPU, and so on.

We can take the lead in the AI era because we are lucky that when we do image processing acceleration, we have accumulated a full set of process experience of "hardware + software + algorithms", from computer graphics to scientific computing.

When deep learning occurs, we are very good at solving this problem.

4. There are a large number of applications in every industry, and if GPU can be used in so many different industries, the demand scenario will be greatly increased.

The reason why we want to be a general GPU is because we think from the developer's point of view.

We hope that the software developed by our ecological developers can be used in the largest user base without having to worry about whether the software can run (so that developers can make more profits, and our ecology will become bigger and bigger).

5.Over the past decade, for researchers in different industries,CPU Moore's Law increases the computing speed by about 100 times, but，Because of GPU and deep learning, scientists' computing performance has been improved by 1 million times.

In the next 10 years, I believe that the improvement of superimposed "algorithm + software + hardware" can be accelerated by 1 million times in some areas of scientific discovery.

6. I thinkArtificial intelligence is the most powerful force, and one of its benefits will be to democratize computer science.

Within our circle of abilities, find the most difficult but meaningful problems in the world and solve them.

7. What we want to do is, within our circle of abilities, to find the most difficult but meaningful problems in the world, and then solve them.

There are two aspects that I am very excited about:

The first is the application of artificial intelligence in various industries. Nowadays, the technology of intelligent voice and image recognition has been mature and can be transformed into applications that produce great value.

Second, this is the next era of artificial intelligence, artificial intelligence must learn the laws of physics.

8. We only provide application frameworks, not consumer AI applications, we just provide future developers with tools, frameworks, and systems to help them build applications in their respective industries.

9. Now there are many cutting-edge machine learning worth paying attention to, and I am paying attention to it.A self-supervised machine learning method for multimodal machine learning (multimodal representation of image, sound, language, etc.).This will take computer awareness to a whole new level, and I'm very excited about it.

Zero shooting learning (zeroshotlearning) and image neural network (Graphneuralnetworks) are also concerned.

I am also interested in the virtual physical world. I think most of the future innovation, manufacturing design, will be realized in the virtual world.

Because the virtual world is in accordance with the laws of physics, and it will be more convenient to do experiments in the virtual world.

10. I think AI's data-driven will become the fourth pillar of science:

The first is similar to Einstein, who found some principles through deduction.

The second is through experimental methods, similar to Newton's method.

The third is through computer simulation.

The fourth, the data-driven scientific discovery brought about by the AI era, will combine the previous methods to bring new innovation.

At present, there is a revolution in computer science, the ability to create the most valuable commodity in the world, and that is intelligence.

11.What is happening now is a revolution in computer science.

Artificial intelligence, a new form of software, is written by the machine itself, which no one has thought of before, which makes every aspect of computing face challenges.

Therefore, the computer needs to collect more data as efficiently and cheaply as possible, which is completely different from the previous workflow.

12. We think that an intelligent car is a computer that can drive itself.

As you know, computers are software-defined, today's mobile phones are software-defined, and computers are software-defined.

These software-defined tools have brought about a technological revolution because software can iterate and implement a variety of functions.

NVIDIA Corp is a computer architecture company, a software company, we are concerned about computer architecture, what we care about is that these software can run on different computers of our architecture.

13. NVIDIA Corp's business serves many industries, so we need a universal truth, and then develop our business around this truth.

Seven or eight years ago, we discovered a universal truth, the first principle.

In the future, all software will be data-driven, and the software will be written by AI, because people can't write such complex software.

So, this means:

We need to have a data strategy, how to capture data, use data, fuse data, and use it for model training.

We needed a whole new infrastructure to meet the computing needs of AI, so we started building supercomputers.

We needed a whole new way to simulate and train AI, so we built Omniverse.

14. The ability to create the most valuable commodity in the world that has never existed in human history is intelligence.

We now have the structure of a model, the structure of a computer science program called deep neural network, which has great expansibility.

This is not Moore's Law doubling every two years, it doubles every six months, the doubling rate is extremely fast, and the comprehensive impact on calculation is incredible.

These computer-created software are expanding and growing at an amazing rate, and achieving amazing achievements.

Our company is continuing to push forward the computers needed for this process, which is NVIDIA Corp's real purpose-to produce intelligence.

15. What is intelligence? Intelligence is the ability to identify patterns, identify relationships, reason and make predictions or planned actions. This is intelligence.

Intelligence is just solving problems.

Enjoy knowing what's going on in the company, which is enough for you to listen to the march of the whole industry.

16. Whenever I solve a riddle or learn something new, it makes me excited.

As the CEO of a technology company, you need to really enjoy and understand what's going on in the company, which is enough for you to listen to the march of the entire industry.

17. We have a basic research department with more than 200 people all over the world.

The organization is incredibly productive, conducting basic research on the entire computing system, from chips, circuits, networks, parallel processor architectures to programming models, from computer graphics to in-depth learning.

They work with all product teams to drive the development of AI computing.

For example, our latest ray tracing RTX is the result of their 10 years of research, which led to NVIDIA Corp's ray tracing technology and optical technology.

18.We want to create supercomputers suitable for AI, and then through our products, we can solve problems that ordinary computers can't solve.

We are building computers for Leonardo da Vinci and Einstein of our time to discover new knowledge and solve problems that would otherwise be impossible to solve.

19. We avoid competing with others, and if others can do it, we won't waste time doing it, which will waste the lives of outstanding computer scientists who work for us.

20. Outstanding people only do jobs that others have not done and are difficult to do. These outstanding people are unwilling to repeat what is already on the market, which will waste their lives.

Therefore, we have created conditions for outstanding talents to come to our company and provide them with a platform for them to complete their life-long work.

I am extremely proud of the work we have done so far.