Nvidia’s most recent acquisition of Arm is the largest since its acquisition of Mellanox in a $7 billion deal. The deal, once completed, is significant for both companies, as it brings together Nvidia’s GPU prowess with Arm’s CPU Power, opening up new opportunities for Nvidia. Nvidia’s top-performing graphics cards allow it to dominate the market with more than 80% market share, while Arm’s customizable and energy-efficient chips could not only give Nvidia access to the smartphone and IoT markets, but also drive its fastest-growing data center Segmentation and AV.
Nvidia acquires Arm
In September last year, Nvidia announced that it would accept the acquisition of chip designer Arm from SoftBank Group for $40 billion. Preliminary terms of the deal include paying parent company SoftBank $21.5 billion in Nvidia stock and $12 billion in cash. Management expects the transaction to close in the first quarter or 18 months of 2022. This does not include any regulatory hurdles, the main source of which may be China. In China, ARM has a huge business base, contributing 20% of its revenue. Additionally, it must be licensed by US and UK regulators. Those regulators have reportedly begun investigating the deal. But if the deal goes through, it will see the rise of a larger supercomputing giant that will bring together Nvidia’s leading AI computing platform and Arm’s vast ecosystem.
Nvidia’s strength lies in its core chipset technology. It specializes in making GPUs, which are well-known accelerators that render images for applications such as games. Among other things, it is optimized for AI that is trained at large batches and at high speed. Nvidia’s competitive advantage lies in its long-term focus on product and software development, allowing it to dominate the market with a market share of over 80 percent, which puts them ahead of rival AMD and others. Last year, Nvidia’s GPU division had revenue of nearly $10 billion.
Although GPUs are capable of handling resource-intensive inference tasks in the cloud and servers, they are not cost-effective for mobile and IoT applications. CPUs have performance, cost and power efficiency advantages. Arm is a professional CPU design company whose products can compete with Intel and AMD’s x86-based processors, which dominate the PC and server markets with a 95 percent market share. That said, Arm-based designs are gaining traction and are already doing well in IoT, edge computing, and smartphones.
So Nvidia has strong coverage in GPUs as a companion to x86 or Arm-based CPUs. The combination of the two companies presents a huge opportunity to combine Arm’s broad CPU portfolio with Nvidia’s GPU and AI computing capabilities for cloud computing, AV, IoT and AI capabilities.
ARM’s processor advantages
The main advantage of Arm processors over x86-based chips is that they have lower power consumption while maintaining high processing performance. Therefore, they are ideal for energy efficient applications such as smartphone application processors. Arm chips already dominate the smartphone market, holding more than 95 percent of the market among major companies including Apple, Samsung and Qualcomm.
In addition to application processors, energy-efficient Arm processors are well-suited for IoT applications, where they have over 90 percent share and are expected to continue to benefit from billions of connected devices extending to edge computing. As such, the acquisition will give Nvidia a significant presence in these markets.
Another distinguishing feature of Arm-based chips is that they are highly customizable and form a decentralized technology ecosystem. Arm has a very unique business model compared to other semiconductor companies in which ARM licenses its IP to other companies to develop chips based on the Cortex family, either through CXC programming partnerships or through custom architecture licensing to build custom CPU cores. Arm, in turn, charges royalties, about 10 cents a piece.
While that might not seem like a huge amount, Arm’s partners collectively ship more than 20 billion chips a year, generating billions in licensing revenue for the company. Under Nvidia’s leadership, management could allow Arm to continue its open licensing model, which has seen hundreds of licensees cumulatively ship more than 180 billion chips.
These highly versatile chips enable a wide range of applications. In addition to smartphones and the Internet of Things, Arm chips are increasingly used in the data center, networking and automotive markets. These emerging areas are critical to Nvidia and are expected to drive shipments and revenue growth, which has slowed over the past few years as the smartphone market stagnated.
In the data center market, it competes with Intel and AMD’s x86-based processors, which have about 95 percent of the market, largely thanks to x86’s better performance, although Arm-based chips seem to be catching up. At the same time, it has the advantages of energy efficiency. Companies like Amazon with Gravitron and Ampere with Arm technology Altra processors are competing with Intel and AMD, and they have also increased Arm’s server market share by 5%.
In addition to this, the automotive market is another key area for the growth of audiovisual products. Arm also belongs to Nvidia’s Autonomous Vehicle Computing Consortium (AVCC). The group consists of leading chipmakers and Tier 1 suppliers that are also Arm partners, such as NXP, Renesas and Harman, working together on a common set of standards for AV deployment. So we think Arm-based chips are very important to be part of that alliance. Overall, we expect trends around IoT, cloud and AV to lead to 6.3% annual growth in shipments and revenue, in line with market estimates.
Nvidia integration
Nvidia’s specialization of GPUs is critical for rendering images and videos in PCs, data centers, and workstations. Arm has blueprints for CPUs with their Cortex line of products, and they also have a line of GPUs called Mali. The merger will give Nvidia the opportunity to integrate Arm’s CPUs with its GPU technology. Its greatest potential lies in the data center.
In my opinion, the company can create an ecosystem of server chips that boost performance without using additional power to enhance its fastest-growing data center segment. Given Nvidia’s presence in markets where Arm has a broad reach, we think this integration could benefit it in consumer devices and automotive as well. Based on Nvidia’s presentation, management believes the integration of Arm will provide a $250 billion target market in 2023.
To calculate Nvidia’s potential synergies, we first estimated Arm’s potential growth based on the data center and server, consumer devices, and automotive segments. Referring to the data provided by SoftBank’s Arm presentation, we calculated the target market forecast to 2025 and obtained the CAGR for each segment. Our estimates show that PCs and workstations will have the highest CAGR at 8.2%, followed by data centers and servers at 4.8% and automotive at 3.6%.
The following illustrates how we see Arm’s integration with Nvidia and the growth rate of the overall market, i.e. how Nvidia’s revenue will grow compared to Arm:
Gaming: Nvidia continues to lead the gaming GPU market with its Turing-based GeForce RTX series, surpassing rival AMD’s over 80 percent share. While gaming CPUs were led by Intel and AMD, ARM gradually emerged in the PC CPU space, especially with Apple’s adoption of Arm-based chips. Still, Nvidia may integrate its products into the high-performance CPU market in the future, but it will take some time. Additionally, it may open up the smartphone gaming market to Nvidia for Adreno chips. According to estimates, gaming GPUs are expected to grow at a CAGR of 29%. Furthermore, the integration of Arm expanded its addressable market by 8.2%. This brings a growth forecast of 39.6%.
Professional visualization: This segment continues to excel with its NVIDIA RTX platform and Quadro RTX series, enabling developers to accelerate and transform their workflows. As the development of virtual reality technology in medicine, architecture and product design is not limited to gaming, the company may consider integrating Arm technology and is expected to exceed market expectations with a compound annual growth rate of 9.8%. Integration with Arm’s workstation GPUs and Arm’s workstation CPUs could grow a further 8.2%, bringing its total growth rate to 18.8%.
Data Center: This segment is growing the fastest. Nvidia could combine its GPUs with Arm CPUs to provide a complete stack for a wide range of uses. Its share of the new TOP500 list of the fastest supercomputers in the world shows that Nvidia continued to hold a 41% share in 2019. Its Tesla GPU family remains strong in HPC and AI inference applications that require more performance than CPUs. As Arm challenges the dominant x86 market, Nvidia could give its super cloud customers such as Amazon, Microsoft and Google a boost. With the market forecasting a CAGR of 49.5%, Arm integration could enable it to grow 4.8% for a total growth rate of 56.7%.
Automotive: As part of the AVCC, NVIDIA continues to build on its leadership with its high-performance NVIDIA DRIVE computing platform. Its SoC powers DRIVE AutoPilot, providing Level 2+ autonomous driving solutions. Potential ways to integrate with Arm to integrate its technology could further enhance its AV capabilities. Without Arm, we assume a CAGR of 33.6% for revenue based on the automotive GPU market forecast. Arm integration with processors for AV can increase the segment growth rate by 3.6%, with a projected total growth rate of 38.5%.
Overall, the combination of the two companies bodes well for Nvidia’s top line growth, including Arm’s revenue and synergies from the integration. Earlier, we expected Arm’s revenue to grow 6.3% in fiscal 2022, in line with shipment growth, which should help Nvidia’s top line. On top of that, in terms of synergies, we estimate the total value in 2022 to be $1.316 billion. In the long run, it will generate $12.717 billion in synergies with greater integration in the PC and data center markets.
If the deal closes in FY2022, Nvidia’s combined revenue with Arm’s sales and synergies is expected to reach $24.368 billion, compared to $21.034 billion without Arm, which is a 57% growth rate compared to 35 percent without Arm. %, an increase of 22%.
Will the deal go through?
If regulators in multiple countries decide to block the deal, the acquisition will fail. Both companies are critical to chip supply and could file antitrust charges. The transaction is expected to close by the first quarter of 2022 at the latest.
However, regulatory threats from the US, UK and China could hinder and delay that timeline. Recent reports suggest that British regulators are closely watching the deal, as well as the risk of Chinese non-approval. In China, ARM owns a joint venture and gets 20% of its revenue from it. The Chinese government is already taking strict steps to keep its chip supply safe from U.S. threats.
If the acquisition does not go through, our revenue growth assumptions will be affected. This will lead to revisions to valuations and price targets based on Nvidia’s revenue without Arm, which is 20% lower than Nvidia’s revenue scenario on Arm.
Paul Santos of Arquiconsult recently gave a good example of why an acquisition of Arm might not work out. His predictions are summarized as follows:
China won’t approve the deal: Trade war tensions escalate and Huawei is added to an entity list that includes the supply of software designed by foreign companies using U.S. semiconductor manufacturing equipment;
The UK will want to keep Arm: The UK will prioritize its local jobs, technical expertise and intellectual property, and prevent its global technology companies from falling into the hands of US companies;
Europe will not approve the deal: Europe will want to block the control of American companies (Nvidia) in order to protect their companies’ access to advanced technologies, to serve the Chinese market and to maintain sovereignty and independence;
While these are good points, we want to argue why this deal is likely to happen.
As far as the semiconductor giant is consolidating market share, the deal doesn’t seem to prove it. While Nvidia has a large share of the GPU market, the ARM architecture will allow it to enter the CPU market, which has long been dominated by Intel and AMD. The two companies still control more than 90 percent of the market.
“Regulators want to make sure their markets have a competitive advantage, which is good for innovation…and that’s good for customers. We can prove it, and we can prove it overwhelmingly, so I don’t have any concerns.” – Nvidia CEO Jen-Hsun Huang said,
We agree that this deal will make the industry more competitive and hope that the regulators will only realize this after further scrutiny.
The world’s largest semiconductor companies are all based in the United States, but the world’s largest semiconductor market is China. The global semiconductor industry is built on this interdependent relationship between the world’s two strongest economies. When “essential products” like semiconductors are needed to drive the advancement and development of global technology, trade laws and political disputes must give way.
This is especially true at a time when the world is shifting to a digital and online environment that relies on semiconductors. If 20% of Arm’s revenue comes from China, it will lose billions of dollars if Arm is prevented from selling to China. Nvidia itself may not push the deal without a deal that would allow them to continue selling to China. We know this is very likely and continue to supply Huawei. On the Chinese side, they also recently approved Cisco Systems’ acquisition of Acacia and stipulated conditions for continued supply to China.
SoftBank and Nvidia are both seasoned players in the M&A game. SoftBank is a private equity firm that regularly engages in mergers and acquisitions as part of its day-to-day business activities. As Japan’s third-largest company by market capitalization, it has been involved in hundreds of buying and selling of companies in the past few years alone. Their wealth or M&A experience and personnel may ensure smooth discussions with regulators and other interested parties.
Nvidia is also no stranger to acquisitions. In 2020 alone, the company completed its $6.9 billion acquisition of Mellanox, a pioneer in InfiniBand interconnect technology used in leading supercomputers and hyperscale data centers. The deal has yet to be done because of the trade war between the U.S. and China. However, 13 months later, regulators in China, the US and Europe approved the deal unconditionally. Nvidia will seek to achieve the same outcome this time around, and now has the experience of an M&A deal with Mellanox.
in conclusion
We think the acquisition of Arm is extremely beneficial to Nvidia and has many opportunities to integrate into all segments. Nvidia dominates the dedicated GPU market with its superior technology and software, with over 80% market share. Arm is a successful chip design company that has sold huge numbers of chips around the world. The combination of Arm and Nvidia would allow Nvidia to combine its GPU-capable competitive advantage with Arm processor-capable GPUs to create a complete stack that would provide even greater competition to Intel and AMD.
Arm chips have become popular in smartphones and IoT devices thanks to their advantages in energy efficiency and design flexibility, but are also seeing increasing use in consumer devices, data centers, and automobiles. These segments intersect with Nvidia’s fastest-growing divisions, giving it the opportunity to integrate its unique capabilities into cloud, edge, AI and AV.
The Links: SKK 323/16 E LQ104S1LG81 BSM50GD170DL