Categories
科技報導

What does Nvidia want to do to acquire Arm? Qualcomm’s shadow reappears with China’s approval



When SoftBank spent $32 billion to acquire Arm in 2016, Nvidia’s market value was still around $40 billion, and its stock price was still around $50. The two companies are still in the same weight class, Huang Renxun appeared everywhere in leather jackets.Four years have passed, Arm is still that Arm, and Huang Renxun is still that iconic leather jacket, but Nvidia is no longer what it used to be.

What does Nvidia want to do to acquire Arm?  Qualcomm's shadow reappears with China's approval 1

Sina Technology Zheng Jun from Silicon Valley, USA

Relying on GPU’s market prospects in many areas such as games, artificial intelligence, data centers, and autonomous driving, Nvidia’s stock price has risen tenfold in the past four years, and has now exceeded $500, with a market value of nearly $320 billion, making it the highest market value. American Chip Company.

The capital market values ​​future growth prospects rather than current performance status. In the second quarter of this year, Nvidia’s revenue was US$3.87 billion, a year-on-year increase of 50%; Intel’s revenue was US$19.73 billion, a year-on-year increase of 23%; Qualcomm’s revenue was US$4.9 billion, due to a significant decline in the epidemic. But the market value of Nvidia is in the US$300 billion level, while Intel and Qualcomm, the two traditional chip giants, are only on the order of US$200 billion and US$120 billion. Nvidia’s price-earnings ratio is as high as 95 times, while Intel is only 9 times.

SoftBank’s acquisition of Arm blood loss

The acquisition of Arm for $40 billion is really nothing to the current boss. According to the acquisition agreement reached by Nvidia, SoftBank and Arm, Nvidia plans to pay $21.5 billion worth of stock and $12 billion in cash to obtain more than 92% of Arm. In addition, if Arm meets the set performance targets, Nvidia will also pay up to $5 billion. And in order to retain talent, Nvidia will also pay $1.5 billion to Arm employees. SoftBank currently holds 75% of Arm, and SoftBank’s Vision Fund holds the remaining 25%.

Arm has not changed much in the past four years under SoftBank, and still plays the role of technology patent provider in the field of mobile chips. Their industry positioning also determines that Arm cannot have significant changes. Arm needs to provide technical design to many chip companies non-discriminatory, just like Switzerland, a permanent neutral country in the field of technology. If you become a vassal of a superpower, give priority to ensuring the interests of certain giants, and interfere with market competition in the mobile field, then Arm will sooner or later lose its current industry influence.

Arm’s performance is indeed not that great. Arm, headquartered in the United Kingdom, does not produce its own chips and does not participate in the competition between chips and terminals in the mobile industry. It only provides architecture design to mobile chip companies. Arm’s revenue includes two parts, pre-license fees and chip patent fees. In the four years since it was acquired by SoftBank, as the mobile industry has gradually become saturated, Arm’s revenue has not increased significantly, and its annual revenue has never exceeded US$2 billion. Due to the decline in shipments of the smartphone industry due to the epidemic, Arm’s performance may decline this year. Perhaps with Apple and Qualcomm gradually launching Arm-based PC chips, Arm’s performance still has room for improvement, but it is difficult to make rapid progress.

This may be why SoftBank wants to sell Arm. Due to a series of investment failures, SoftBank needs to adjust its investment portfolio. When Sun Zhengyi acquired Arm, he said that this was the most important acquisition transaction in his life. But from the perspective of return on investment, investing in Arm is another blood loss for Sun Zhengyi. SoftBank has held Arm for four years and only gained $8 billion in appreciation. But in order to raise funds for the acquisition of Arm, Sun Zhengyi did not hesitate to reduce his holdings of Alibaba stock to cash out 10 billion US dollars. At that time, Ali’s stock price was still around US$80, and now it is above US$270. In addition, Sun Zhengyi sold the game company Supercell to Tencent to raise US$8.6 billion.

What does Nvidia want to do to acquire Arm?  Qualcomm's shadow reappears with China's approval 2

Build an AI super giant

For SoftBank, Arm is just a tasteless investment asset with little room for appreciation. But for Huang Renxun’s Nvidia, Arm’s strategic value is far more than 40 billion US dollars. Arm’s important position is not measurable by performance. At present, more than 90% of the world’s smart phone chips use Arm’s architecture. Whether it is Apple or Qualcomm, or Samsung, Huawei, or MediaTek, it is temporarily inseparable from Arm. Nvidia has completed the acquisition of Arm, and has mastered the foundation of global smart phone chips.

What does NVIDIA want to do after getting Arm? Huang Renxun clearly stated in the open letter of acquisition that he will not change Arm’s current open authorization model and maintain its global customer neutrality. Arm will continue to be a subsidiary of NVIDIA, and its headquarters will remain in Cambridge, England. But he plans to apply Nvidia’s technology to the Arm ecosystem and set up an AI R&D center in the UK.

Although many hardware and software giants in the smartphone industry are already focusing on AI functions, Apple, Qualcomm, and Huawei have all introduced processors that emphasize AI computing performance, but if NVIDIA penetrates its AI technology into the Arm chip design, It is equivalent to completely changing the current layout of the AI ​​industry from the underlying architecture and reversing the embarrassment of Tegra’s failure in the smartphone industry. Even if NVIDIA’s Arm continues to sell technology patents, NVIDIA’s AI technology can also cover the entire smartphone industry in the future.

Nvidia once had a smartphone processor business, and there may be an Nvidia version of Xiaomi Mi 3 from the outside world. But because the technical problem could not be solved, the market was not recognized, and was eventually forced to give up. Tegra chips are still used on Nintendo’s Swith game consoles. Obviously, behind the acquisition of Arm is Nvidia’s ambition to expand again in the mobile field. They hope to combine their AI technology with Arm chip design to bring an impact to the mobile computing field.

Intel is a direct sufferer

The real purpose of Nvidia’s acquisition of Arm is to push its AI strategy to new heights, especially to open up the broad market for cloud computing. According to Huang Renxun’s plan, the future Arm chip design will run through AI technology, which will not only directly affect the smart phone industry, but also the data center market. Perhaps the most direct impact of this transaction is Intel. The data center is now Nvidia’s largest business, and revenue has exceeded the gaming business.

Compared with the already saturated smartphone industry, the cloud computing market with huge growth potential may be an area that Nvidia values ​​more. Huang Renxun also made no secret of his ambitions in this field. “Data center and cloud computing markets are looking forward to Arm processors. Power consumption will directly affect computing performance, computing throughput, and configuration service costs.” Although the current data center market is still dominated by Intel x86 architecture, it has low power Consumption is increasingly becoming a key factor in data centers, and many manufacturers are already attacking this field based on the Arm architecture.

Here is an interesting gossip. After the departure of former Intel president Renee James, he founded a chip startup Armere. Their business is to build server chips based on the Arm architecture, and they have also received Arm’s investment. Zhan Ruini was an assistant to Intel legend Andy Grove. She was also a candidate for Intel’s CEO, but left before Ke Zaiqi resigned due to the scandal.

In the past two years, Amazon, Qualcomm, and Huawei have all launched their own Arm-based products in this area, and Google has also launched its own TPU processor. And with the maturity of machine learning technology, the data center processors required for cloud computing in the future will gradually shift from the previous multi-core CPU to parallel processing and coexistence of different types of processors, including the GPU at the core of NVIDIA. If Nvidia relies on Arm plus AI plus GPU to achieve breakthroughs in the field of server processors in the future, the value brought from this market alone will be far more than 40 billion US dollars.

Antitrust approval is the key

However, before realizing his own AI penetration Arm’s strategic vision, Huang Renxun still needs to face a more difficult challenge: anti-monopoly approval. The giants that “want” to acquire Arm are far more than NVIDIA, but giants such as Apple, Intel or Qualcomm are very clear that they “cannot” acquire Arm because their business overlaps with Arm and cannot pass antitrust approval. . The reason why Nvidia finally made the move is also because they currently have almost no smart phone, flat panel and TV processor business.

Although Huang Renxun has repeatedly emphasized that he does not intend to change the industry positioning of Arm technology licensing in the short term, this does not alleviate the industry’s worries about this transaction. Apple, Intel, Qualcomm, Huawei, Samsung, MediaTek, and Spreadtrum will closely follow this transaction. They are even more direct competitors of Nvidia in business areas such as future automobiles, artificial intelligence and data centers. These companies may push the US and Chinese regulatory authorities to strictly review the Nvidia transaction.

National strategy is also unavoidable for Nvidia. After Nvidia acquired Arm, Arm became a subsidiary of the American company. Not only Chinese regulators will take this factor into consideration, even European regulators will consider it carefully. Arm co-founder Hermann Hauser publicly stated that the trade would be “a disaster that will destroy its business model.” Hauser said, “Arm is the last European technology company with global relevance and it was sold to Americans. The deal announced overnight will destroy ARM’s business model as the “Switzerland of the semiconductor industry”, and NVIDIA It is a competitive relationship with ARM’s customers.”

Huang Renxun stated to the Chinese media yesterday that he was optimistic about the prospects of China’s approval. “Chinese regulators will welcome this transaction, because NVIDIA will bring more solutions to the market, which will benefit Chinese customers and the market.” His optimistic statement does not mean much. A few years ago, Qualcomm CEO Molenkov was better than Huang Renxun. More optimistic, but gave up sadly.

Qualcomm abandoned the shadow to reappear

From October 2016 to July 2018, Qualcomm spent 38 billion U.S. dollars to acquire NXP Semiconductors (NXP) in the Netherlands. After a long wait of 20 months, it finally announced its abandonment. Qualcomm paid NXP a breakup fee of up to US$2 billion for this. The reason for not completing the acquisition is simple. Among the nine major market regulators in the world (the United States, the European Union, South Korea, Japan, Russia, etc.), only China has not approved the transaction.

The two transactions have many similarities. When Qualcomm acquired NXP, it did not have much business in the field of automotive chips, and the businesses of the two companies did not overlap. Qualcomm chips focus on mobile computing and communications, while NXP chips focus on smart cars, Internet of Things, mobile payments and security. After Qualcomm acquired NXP, it is equivalent to spanning many computing industries. Although Qualcomm promised to continue to authorize NXP technology in the next eight years and the authorization standards will remain unchanged, in the end, it did not wait for the green light of the Chinese regulatory authorities.

Why did China delay approval of Qualcomm’s acquisition of NXP? The possible adverse impact of this transaction on competition in the Chinese industry is certainly a consideration, but national strategy is also another key factor. 2017-2018 coincided with the political environment of Sino-US trade negotiations and tariff clashes. The United States once again imposed an export ban on ZTE. After the Chinese government intervened and ZTE paid a second fine of US$1.4 billion, it was released.

However, in 2018, the United States had not resorted to the “physical list” of its killer tricks against Huawei. It only prohibited small US operators from purchasing Huawei and ZTE products. This is not a major blow to Huawei. The United States officially banned Huawei from 2019 and banned Huawei from using American components and technology; and implemented a second wave of bans this year, severing cooperation and procurement between Huawei and non-US companies such as TSMC, MediaTek, and Samsung. Arm must also comply with the US ban.

The current environment is obviously worse than 2018. After Nvidia acquired Arm, it is equivalent to the patent foundation of China’s smartphone industry, all pinched in the hands of American companies. Merely committing to continue to open up authorization has no practical meaning. Nvidia is an American company that must comply with various orders of the US government. Considering a series of developments this year, Nvidia will obviously have to make greater efforts and concessions than Qualcomm for the Chinese regulatory authorities to agree to the acquisition of Arm. At least for now, the prospects for transaction approvals are hardly optimistic. Nvidia estimates that the approval process may take up to 18 months.

What does Nvidia want to do to acquire Arm?  Qualcomm's shadow reappears with China's approval 3

Another tattoo?

An interesting detail is that the details of the break-up fee have not been announced for this transaction. The outside world is currently unable to know how much liquidated damage NVIDIA will pay to SoftBank if the transaction is abandoned due to antitrust approval. By the way, Huang Renxun has an Nvidia logo tattoo on his upper left arm. It was tattooed when he celebrated that Nvidia’s share price exceeded $100. If the successful acquisition of Arm to achieve the AI ​​grand strategy pushes the stock price to more than $1,000, maybe boss Huang will celebrate again on his right arm. But he needs the approval of the Ministry of Commerce of China.