Categories
科技報導

Huang Renxun appeared to answer questions: ARM China will be part of NVIDIA



Fortunately, we are indeed witnessing a major historical moment in semiconductors. Yesterday afternoon, we participated in the online media Q&A session of “Nvidia Acquired ARM”, and Nvidia CEO Huang Renxun and ARM CEO Simon Sagars (Simon Sagar) attended this Q&A session. The questions in the minds of developers and even the general public in various countries were conveyed to the two key figures in this transaction through the mouth of the media.

Produced | Tiger Sniff Technology Group

Combing | Xiong Guang’s Friends

Huang Renxun appeared to answer questions: ARM China will be part of NVIDIA 1

We have selected and compiled all the answers of Huang Renxun and Situkes. Different people have different opinions whether they have answered your questions.

Huang Renxun:

I am happy to announce that this transaction has been agreed by both parties and it will be one of the greatest technology companies in the history of the world. For the technology industry and the British government, this transaction is of great significance. We will jointly build the world’s premier computing company. Before closing the deal, let me explain how we view the world and how this deal will fit into it.

AI is the most powerful technical force in the current era. Related software processes large amounts of data by using supercomputer-level processing capabilities. AI can create models to accurately infer intentions, predict upcoming actions or understand complex relationships.

AI capabilities are moving towards every corner, airport shops, warehouses, and hospital streets. AI is moving from the cloud to the terminal, where smart sensors connected to the AI ​​computer can check video clips, coordinate traffic, and save power.

Over time, there will be trillions of small autonomous computers powered by AI, connected to powerful cloud data centers in every corner of the world, and these small computers will continue to make smarter predictions and faster responses.

AI computing, network and data processing are more important than ever, and Nvidia and ARM are the leading computing companies in this area.

This is the background to why this transaction is so important to the two companies.

And ARM exists in the vast ecosystem of terminal devices, with more than 180 billion units. By merging our companies, we will be able to expand and visualize AI computing platforms to arm a broad ecosystem. Together, we will create next-generation platforms for global computer manufacturers, from cloud data centers to mobile phones, to driverless cars, to the Internet of Things.

This combination is financially meaningful. We expect to immediately increase gross profit margin, as well as the creativity of earnings per share and closing price.

But what we can get is much more than that.

For our employees, this will mean new opportunities and career paths to work in one of the fastest growing technology companies in the world. For Nvidia, it will become the cornerstone to promote the development of our developer ecosystem or our computing platform. The combined company will provide new innovation to the customers of the two companies. At the same time, we will maintain an open licensing model and customer neutrality to provide services to customers in any industry worldwide.

This is critical to Arm’s long-term success. We like the cooperative business model.

In fact, we intend to expand Arm’s licensing portfolio through the use of proprietary technology. This combination will enrich our two ecosystems. In recent weeks, some media have speculated about what this deal might mean for Arm and Arm in the United Kingdom and the United States, so let me clarify some things.

After the transaction is completed, we hope to develop the Arm ecosystem and help it become bigger. The reason we want to acquire Arm is because it is an amazing company and has hired some of the world’s greatest engineering experts. There is no doubt that we want more great engineers.

We hope that this work can be completed in Cambridge, England, where Arm’s business has begun and is continuing. There are intellectual property created by thousands of outstanding engineers over the years, and Arm will remain in the UK in the future.

But most importantly, we want to further improve our level. We hope to promote Arm to the global AI leadership.

To help achieve this goal, we will create a world-class AI laboratory in Cambridge, where we will build advanced supercomputers, where leading scientists, engineers, and researchers in the UK and around the world can work in healthcare, self-driving cars Breakthrough work in other fields.

In addition, we hope that the UK will attract the best talents from all over the world. We have established a cooperative relationship with King’s College London and the NHS medical system in the United Kingdom. We will open an AI R&D center in the United Kingdom. Our opportunities in technology have never been more exciting.

ARM CEO Segal:

Good morning, everyone. We are very excited about the upcoming prospects. In the past 30 years, we have established a global chip standard at the Cambridge headquarters and expanded our product portfolio through substantial investment. Now, we are standing in front of the opportunities of the new computing era. In our history, we have seen that the number of embedded computing applications is growing.

We have provided support to the world semiconductor industry through our technology. Every day, we see new use cases. However, the definition of computing is constantly evolving and changing. Now, we see an advanced world defined by AI, and we see that AI is entering the stage of mature applications, where more and more applications can be found every day to start using AI.

Now, by combining with Nvidia, we believe that we can continue another new journey as we did in the past. We can help the entire world semiconductor industry realize this vision and develop chips with embedded artificial intelligence functions.

We believe that there will be a very exciting future ahead.

Question 1: When SoftBank acquired ARM in 2016, the British government made a legally binding commitment to the number of employees and established its headquarters in Cambridge. So, can Simon answer, can you provide any legally binding commitments for this transaction? If this is not the case, it fails. Have you given a specific informal commitment?

The second question is some actions of the US authorities. Do you foresee any related problems as you continue this business?

A: First of all, we openly discuss with the British government to establish a partnership so that we can invest in the UK and expand the business of the R&D department. It is very important that we establish a world-class technology research center in the UK, where we will develop one of the most powerful supercomputers in Europe.

We hope to attract world-class researchers from the UK to engage in groundbreaking work. These will all take place in Cambridge. I believe that the British government will realize that we are making a very significant investment in the UK, so it will definitely have a dialogue with them. We have an implementation framework, but no binding commitments.

Follow-up: Have you made an informal commitment in writing to the work you will do, but the plan has not yet started?

We have just started a dialogue with the British government. We have enough time to do this and we will do it. Another problem is that as far as technical jurisdiction is concerned, the relevant content of the transaction has not changed much from before. The technology originated in the UK, and it will remain in the UK because the department will still be headquartered in the UK and will continue to work in the UK.

Question: ARM’s business model is to sell IP cores to many chip design companies, so independence must be maintained. And four years ago, ARM said by itself that it has been completely independent since its IPO, which is what our partners attach the most importance to.

But changes that may occur in the future may make ARM no longer need to be independent. So will this change cause any concerns from major customers like Apple or Qualcomm? How do you adjust the business model that ARM has had for decades?

Answer: Independence is absolutely part of our strength and absolutely part of all values. The value of ARM lies in how we cherish the effective combination of the technology and business model we create.

In fact, by merging with Nvidia, we will be able to make greater investments in the technology roadmap than we have now. As part of the merger, we will definitely maintain the existing business model.

Therefore, we may belong to a different company, which is equivalent to changing our parent company. Our goal is still to continue to maintain a certain degree of independence. This means that we can still serve the entire world semiconductor industry through the authorization model.

Huang Renxun added: The advantage of independence is that it will ultimately bring openness and fairness. We definitely have the intention to do so, because maintaining this business model is the best way for us to continue to expand our business scope and achieve success, and this requires us to make great efforts for this.

What Simon and I are looking forward to is that although the future of computing is still very important for mobile devices, there are many types of computing architectures we need. In addition to the cloud, there are high-performance computing edge data centers, robotics, and so on. Although the architecture will be very similar, their shape and size will change. We will add artificial intelligence to them. If this is to be done, we will have to make a substantial investment.

Therefore, I hope to combine the capabilities of our two companies to explore new opportunities faster.

Q: Are you (Huang) willing to make a firm and quantitative commitment to the total number of employees in the UK in 2016, like SoftBank?

Answer: We are a very different company and we can provide very different things. The best way to think is that we have an open mind to establish a constructive framework to reflect our intentions and our investment opportunities in the UK. So we will definitely have such a dialogue, we will do something meaningful.

Q: The two founders of Arm commented that selling Arm to a US company means that they will be supervised by the US Foreign Investment Commission, and the US may eventually ban or restrict the use of Arm licenses by Chinese companies. Can you confirm whether this will happen?

Answer: Any product exported must understand the export control rules of the country/region in which it operates. Therefore, we developed products in the UK, some products in the United States, and products in Europe, India and many parts of the world. When we consider export restrictions, we need to confirm what the rules are and analyze the product itself, which has nothing to do with the ownership of the company.

Therefore, many of our products are designed outside the UK or the US, and most of our products are not affected by US export controls. But rules are something we must be careful about, and we must abide by all regulations surrounding export products. Changing the parent company will not change our requirements for this and the impact on our products.

Q: You (Huang) mentioned that Nvidia’s GPU and AI technologies will become part of Arm’s expanded product portfolio. Can you clarify further which part of your GPU and AI technology will be available in that extenders IP combination? Does this include Mellanox technology?

A: I think the final answer is that we want to provide as many products as possible to our customers. You know, one thing we have been unable to do is to reach the breadth of partners in the ARM ecosystem. As Simon said just now, it is used on a variety of computers to calculate computers of different sizes and shapes. In the near future, artificial intelligence technology will be integrated, so we want to use Nvidia’s technology through this huge network, so that people can create computers of different sizes and shapes.

In the future, the ARM ecosystem will be injected with AI technology, but it ultimately depends on the needs of customers. Just like your observation today, many SoCs today do not let us benefit from use.

We now have Mellanox (another company that Nvidia acquired not long ago), which is the global leader in high-performance computing and high-performance networking technology. By expanding the server connection network and their thousands of partners, we can serve more customers and more markets.

Q: I want to know what kind of pressure this acquisition will put on NVIDIA’s balance sheet? This is a big acquisition. Maybe you (Huang) won’t leave a lot of money next year.

A: The incredible news is that SoftBank has accepted most stock transactions. Mr. Sun is a big fan of Nvidia and a big fan of Arm. We are one of the first two investments of SoftBank’s Vision Fund. Nvidia is the world leader in artificial intelligence computing, and Arm is the most influential computing company in the world. The combination of these two companies allows us to bring artificial intelligence computing into every industry and help create the future of the Internet of Things. This is exactly what he described a few years ago, so he traded in the form of stocks.

Our company is growing very fast, we are one of the fastest growing technology companies in the world, and we are very profitable. This transaction is value-added after the transaction ends, so this is a great deal, it is really a wonderful deal.

Q: How will you lead a completely different team? What is the working method?

A: The integration process will only begin at the moment when this transaction is completed. Before that, we were a completely independent company. Therefore, we have a lot of time to get in touch, and recently we have also communicated with mellanox. In fact, the great thing about this is that the two companies complement each other and do completely different things—the products they produce are different, and the markets and business models they serve are different. There is almost no overlap.

We will look for opportunities to share best practices and learn from each other. However, we also need to maintain the independent nature of the business department. Neutrality is very important for customers.

Question: Will you (Huang) continue to develop Mari GPU cores? The second question is, will you continue to develop Risc-V products?

Answer: The answer to both questions is yes. We are all users of Arm and Risc-V, and all of them are loyal users.

I know they all look like microprocessors, but they are very different things. You can apply the same logic to Nvidia’s GPU computing platform and Mari GPU. They are exactly the same.

In the same analogy, one of them is a computing platform, which has a rich software and tool ecosystem, and many third-party developers, many third-party developers integrate with it and rely on it to build products around it. Arm’s rich ecosystem makes it different from Risc-V.

We use Risc-V as internal cpu and controller in the company, which is the perfect application of microprocessor on large chips. This is also true in GPU computing. The two are very complementary, and we will continue to use both methods at the same time.

Question: Some people think that there is an artificial intelligence competition between China and the United States. I would like to know your opinion? Especially what is the impact on this acquisition?

Answer: Good question. There is actually an artificial intelligence competition all over the world. Obviously, from healthcare to transportation, as well as logistics and manufacturing as a whole, we will see the productivity gains created by new technologies, and the gains in capabilities that winners see from artificial intelligence are very profound.

Every country is racing to establish a computing center, hiring a large number of researchers and engineers who know how to carry out breakthrough work in this field and apply technology to various industries.

This is why we are making this investment in the UK and why it makes sense to combine GPU and ARM technology. Cambridge can also attract other researchers from all over the world. One really important thing is that it is really difficult to attract, retain and operate a company with world-class engineers, and resources are very scarce.

This is why there are few great technology companies in the world.

Speaking of China, our recent acquisition of Mellanox has actually gone through regulatory procedures. We have accumulated some experience in this area, and I think it will be helpful to this transaction. China has a large number of very professional developers and customers.

Q: Arm has a 49% stake in a joint venture in China. Do they have any feedback on this transaction?

Huang A: We have recently been reviewed in China with Mellanox, and we have accumulated some experience in this area. We can bring more solutions to the market, which is very beneficial to Chinese customers.

Arm has a joint venture company in China. The structure of the joint venture company will not change. We will own a minority stake in this joint venture company. Now it will become a part of NVIDIA.

Simon: It’s that simple. Arm currently owns shares in a company we established in China two years ago. This is our way to gain access to the Chinese market. We have a very wide range of customers and technology licenses who develop new applications based on Arm, and we hope this situation continues.

China is an important market for Nvidia. After the transaction is completed, Nvidia will own equity in China.

Simon added: ARM China is a joint venture that we have previously established. In this way, we hand over IP licenses to Chinese customers. There are many excellent engineers there. They do a good job in this area, do outstanding work, develop new applications, and we hope that in the field of artificial intelligence can continue to occupy a place in China’s important market. There will be more possibilities for the future development of this part.

Q: If the transaction may be rejected by the regulatory agencies of other countries, are you (Huang) ready for Plan B?

Answer: Plan B is enough. This opportunity is actually only once in my life, so I seized the opportunity. The combination of a company with leadership in the field of AI computing and a great company with a huge ecosystem, and stepping into computing fields of all sizes and shapes, will create the most efficient computer in history.

I know this is a very lucrative deal. This is very beneficial to SoftBank and also very beneficial to us. I think that regulators usually only support this very much.

Q: How long is the supervision process expected to last?

Answer: Unpredictable. For Mellanox, we hope to complete it within 18 months, and we did it. In this transaction, we think it is reasonable to spend more than a year. We have had a lot of discussions with regulatory agencies and explained in many aspects such as technology and market dynamics. This process will take at least one year.

Q: For Simon, what do you think of the partnership with SoftBank in the past four years? For Huang, how will you build relationships with SoftBank after it becomes a shareholder?

A: Let me talk about my experience as part of Softbank in the past four years. At that time, we needed to make a lot of investment in technology and use this investment to expand the product base and market boundaries. SoftBank has always been a supporter of Arm and has fulfilled their business investment commitments. We have upgraded the technology related to data centers. In terms of 5G, we have expanded the IP product portfolio. Today’s product portfolio is much more than four years ago. Therefore, it is a great experience to be a member of SoftBank.

Huang: Sun Zhengyi is one of the great visionaries. He has gone through the PC revolution and the mobile Internet revolution. Now it is the AI ​​era, that is, the AI ​​revolution. He has made very successful investments at each of these moments. Now, we need to adopt a new computing method that combines AI technology with computing architecture and covers a wide range of industries. Previously, the Internet has affected every industry, but now new computing methods will affect computers of all sizes and shapes around the world.