Why did Nvidia acquire ARM for $40 billion?
Nvidia announced yesterday that it will acquire ARM for US$40 billion, marking the largest M&A transaction in the global semiconductor market to date. People can’t help asking: What exactly does NVIDIA intend to do?
Nvidia (Nvidia) formally reached an agreement with SoftBank Group. According to the agreement, Nvidia will acquire British chip manufacturer ARM from SoftBank Group and SoftBank Vision Fund, with a transaction value of approximately US$40 billion. If the transaction can be successfully completed, in terms of dollar value, this will be the largest semiconductor transaction in history. But in fact, measuring this transaction based solely on the value of the dollar may far underestimate the importance of this transaction.
ARM is the core of almost every smartphone sold today, from Apple iPhones to Samsung Galaxy devices running Google’s Android software. In addition to smartphone processors, ARM technology is also used in many smaller chips integrated into various computer systems and devices. ARM said that its processor cores and other components of the chip have been shipped to all parts of the world, the number of more than 180 billion pieces.
Jensen Huang, the founder and CEO of NVIDIA, told the media on Monday: “They (ARM chips) are everywhere, powering all types of computers, from the world’s fastest supercomputers to smart watches and smart thermostats. .”
But for Nvidia, ARM represents an opportunity to boost its fast-growing business of selling chipsets and software to data centers. Investors believe that Nvidia can closely integrate ARM’s CPU technology with its expertise in graphics processing units (GPU) to establish a leading position in other areas of the semiconductor industry.
Investment research firm Jefferies analyst Mark Lipacis (Mark Lipacis) said that by letting Nvidia create an “ecosystem” of server chips and software, the deal will further boost its data center business. Nvidia can integrate the ARM processor more closely into its own products to improve performance without using additional energy consumption.
In addition, this transaction will further strengthen Nvidia’s position and make it more difficult for competitors to shake its business. Lipasis said: “With ARM, NVIDIA will also provide general-purpose, heterogeneous data center-scale solutions, and ultimately get the data center ecosystem’s serial processing part (serial processing part) up to 80% of the value. In addition, ARM also allows NVIDIA to have a larger “moat” in the data center market.”
This year, Nvidia’s stock price has risen by more than 110%. After the deal was announced, Nvidia’s stock price rose by more than 5% on Monday, indicating that investors are confident in the prospect of the deal.
Sales chip blueprint
Below are some of the angles touched:
Both ARM and Nvidia are “silicon-focused” companies, but they focus on different areas. Huang Renxun said that ARM and Nvidia are “complementary.”
ARM is headquartered in Cambridge, England, and sells blueprints needed by other companies to produce low-power chips to drive devices such as smartphones, tablets, virtual reality (VR) headsets and smart speakers. It is called the instruction set. Some customers, such as Apple, design their own CPU cores after obtaining authorization from ARM. Other companies, such as Qualcomm or Samsung, obtain licenses for the entire processor design directly from ARM.
ARM makes profits by selling intellectual property licenses, software, and tools to chip designers who want to use its technology, and collects royalties on each chip that uses its technology. In 2019, ARM generated more than $1.7 billion in revenue.
ARM has a long history. It was founded in 1990 as a joint venture established by Apple, Icon Computer Corporation of the United Kingdom, and American chip manufacturer VLSI Technology to develop chips for Apple’s personal digital assistant, Newton.
But 15 years later, driven by the smartphone boom, ARM has leapt to the forefront of the industry. Smart phones require chips that consume much less power than x86 notebook and desktop chips produced by Intel. Although their computing power is not as good as laptops, at least the battery life is much longer than laptops.
The first-generation iPhone used a Samsung processor and an ARM core, but its computing power was limited. Since then, chips based on the ARM core have become the standard for Apple and Android phones, and have become significantly more powerful, even comparable to x86 chips in certain tasks. For example, Qualcomm chips used in Samsung, Google and Huawei mobile phones use ARM cores.
In addition, ARM is also entering the market for other types of computers besides smart phones. For example, Apple is planning to switch its Mac notebooks and desktops to its own chips, which are officially based on ARM’s chip technology. Microsoft also spent a lot of time and money to transform Windows to make it run on ARM chips. Today, ARM chips are increasingly being used in data centers because their low power consumption is an obvious advantage.
When SoftBank acquired ARM in 2016, the company also emphasized the potential of the “Internet of Things”, which means that more and more items, such as machinery or cars, can be equipped with Internet connections and a cheap microprocessor. Of course, these microprocessors are all running on ARM technology.
Nvidia was initially known for designing graphics chips, which enable the central processing unit to reduce the burden of 3D graphics work to improve computing power and efficiency. Its customers are mainly gamers, companies that build game systems, and people who need to build 3D graphics.
Games are still a large part of NVIDIA’s business. But in the past ten years, NVIDIA has begun to expand into many different areas.
In 2008, Nvidia began to build its own “System on Chip” (SoC) hardware under the “Tegra” brand, combining its graphics processor and ARM processor core into one product. For example, the Tegra chip is the core of the Nintendo Switch game console, although it is not as attractive as Qualcomm’s Snapdragon processor among smartphone manufacturers.
In addition, NVIDIA’s self-driving car platform also includes hardware chips with ARM cores. Today, Nvidia is starting to manufacture chips and software for self-driving cars and robots.
In fact, in recent years, the biggest positive factor behind Nvidia is the rise of artificial intelligence (AI). Researchers found that compared with old server chips or simple CPUs, Nvidia’s graphics processors are more suitable for running artificial intelligence algorithms.
As Nvidia’s ambitions continue to expand, the company’s wealth also grows. According to data from financial data service provider FactSet, when the market opened on Monday, Nvidia’s market value exceeded 322 billion U.S. dollars, an amazing increase compared to five years ago, when its market value was only about 12.3 billion U.S. dollars. In the 2020 fiscal year ending on January 26 this year, NVIDIA’s revenue was 10.92 billion U.S. dollars, a decrease of 7% from 11.72 billion U.S. dollars in the same period last year.
What will happen in the future?
Some analysts believe that the focus of this transaction is to boost Nvidia’s data center business. Currently, the data center is Nvidia’s fastest growing business. In the quarter ended June, Nvidia’s data center revenue was $1.75 billion, a year-on-year increase of 167%, accounting for about 45% of the company’s sales, and surpassing its gaming division for the first time.
In addition, this transaction also gives Nvidia the opportunity to sell its AI chips and graphics chips to ARM customers. Huang Renxun said: “If we can use their (ARM) business model, if we can stand on the network they created and put our invented technology into this channel, so that their thousands of partners can get it, For us, the economic benefits will be incredible.”
Of course, many of ARM’s customers are competing with Nvidia, including Samsung, Qualcomm, Broadcom and Texas Instruments. These companies may be reluctant to obtain additional technology from competitors. Nvidia stated that it will continue to operate ARM as an independent subsidiary and adopt a “customer-neutral” and “open license” model. If these companies are willing to pay, they can obtain a license, even if they are direct competitors of Nvidia.
Huang Renxun also said that ARM CEO Simon Segars (Simon Segars) has communicated with many customers and answered some of their questions.
Huang Renxun said: “They want to make it clear whether they can continue to rely on ARM. And our ability to announce this deal shows that the negotiations are going well. Although no one likes to change, if they realize, we will continue to work on the current business model. , Continue to adhere to the open and fair policy, there will be no worries. More importantly, we hope to keep everything the same, and will also add Nvidia’s intellectual property rights, so that they can have more things to buy, and we can also They provide more value.”