Challenges Facing the Mobile and Computing Industries with Arm IPO

This week, NVIDIA and SoftBank jointly announced they are abandoning their plans for NVIDIA to acquire Arm Limited from SoftBank, almost eighteen months after the two companies announced the transaction worth US$40 billion. SoftBank was looking to get rid of Arm almost four years after acquiring the British chipset designer in 2016. In 2020, it approached several potential buyers, but NVIDIA managed to make the best offer, including a bulletproof strategy for enabling Arm’s Instruction Set Architectures (ISAs) to compete effectively against x86 compatible processors across multiple industries, including Personal Computers (PCs), High-Performance Computing (HPC), Artificial Intelligence (AI), and the Internet of Things (IoT).

Impact on NVIDIA: One direct impact the termination of this transaction has on NVIDIA is that it will have to pay US$1.25 billion to Arm in accordance with the breakup clause it signed with SoftBank. This is in addition to the legal and marketing fees it spent to incentivize stakeholders about the importance of the deal to the entire industry. On the strategic side, NVIDIA may want to review its options for its Central Processing Unit (CPU) business, and eventually consider RISC-V as a credible option if Arm fails to meet the expected innovation requirements.

Impact on Arm: Financially speaking, Arm will gain a potential US$1.25 billion out of the failure of this deal, which is equivalent to eighteen months’ worth of Arm’s Earnings before Interest, Taxes, Depreciation, and Amortization (EBITDA). This is a serious amount of cash, probably enough to keep the British company’s operations alive and prevent shares from plunging in the next couple of years. However, this cash is not enough for Arm to fulfill its ambitions to lead in AI and HPC, where x86 and NVIDIA dominate. Compare this amount to the annual Research and Development (R&D) budget of Intel, which is around US$13 billion, and a great portion of it would be spent on developing Intel’s core business, notably its x86 ISA. This indicates the financial firepower that will be required to compete in AI and the HPC data center business.

The failure of this transaction will also have some serious consequences on Arm, including the toxic culture inside Arm created from the struggle with the acquisition process. This means the process to an Initial Public Offering (IPO) for Arm may be accelerated to avoid any further churn of Arm’s talent. However, the IPO itself could expose Arm to further challenges, as it could cost hundreds of millions of dollars and take up to two years to materialize. Not only would this put a burden on the company’s financial performance, but it could also expose the company to additional distractions and cultural challenges. IPOs are often driven by short-term gains, so long-term strategies can be compromised. It is likely that Arm would have to prioritize some areas of research over others, which could compromise innovation. So, it is questionable whether an IPO path, even with a generous valuation, would allow Arm’s ambitions to innovate beyond the mobile market and explore new opportunities, notably in the HPC and AI markets.

Impact on the Mobile Industry: The impact of this development will be minor on the mobile industry in the short term but taking Arm to an IPO could have a long-term strategic impact. Mobile phones continue to push the innovation boundaries very high. Arm licensees in this space will continue to push the company to adopt new extensions and features in the next-generation Arm ISAs and Computer Processing Unit (CPU) micro-architectures. The increasing variety of accelerators within the mobile environment and the diversity of their implementations is challenging Arm to maintain or cater to all of these implementations. This issue will become more apparent as mobile devices develop more sophistication and are required to handle advanced AI workloads (at all levels, from Radio Frequency (RF) management to application enhancements), higher resolution video and imaging, and several other applications that require advanced hardware accelerators. If Arm does not accelerate its ISA roadmap to accommodate these innovations and efficiently exploit the next generation of acceleration technology and applications that will use them, the Arm ecosystem could be exposed to huge fragmentation. It is essential that Arm is afforded the resources and opportunity to address the looming fragmentation issues. Again, these efforts require long-term strategic dedication to expand and innovate in line with the industry’s requirements and away from the financial market pressure an IPO could bring. The IPO also presents a potential risk of loss of control by the executive team when it comes to making the right strategic decisions that could benefit all stakeholders' interest and innovation in general. This pressure could constrain innovation quite dramatically, if not managed properly. The worst scenario could lead to the collapse of the company all together. Although this is unlikely, the mobile industry should be prepared for potential disruption induced by Arm’s IPO transition. This could lead to a fragmented environment not only at the hardware level, but for the Operating System (OS) and software development environments as well.

Impact on the Rest of the Industry: The announcement to take Arm to an IPO will create a certain panic within the wider industry, pushing certain industry players to evaluate their “Plan B” in case Arm struggles with its IPO process. For, instance, some players may consider adopting alternative ISAs, such as RISC-V, although ABI Research believes that the RISC-V ecosystem is not yet ready to be able to challenge established ISAs, such as Arm and x86. Some other players may even push Arm to open source its ISA framework and micro-architecture stacks. However, this approach will require a strong governance model, so the interest of the entire industry is guaranteed to be away from competitive gains and strategic ambitions certain companies may have to expand their spheres of influence. The failure of NVIDIA to acquire Arm is also great news for x86 players, notably Intel, which was fearing that this acquisition would lead to a stronger Arm ISA framework. Bear in mind that competing against x86 was core in NVIDIA’s strategy to acquire Arm. Through the acquisition, NVIDIA intended to accelerate the migration of Arm ISAs and micro-architectures toward heterogenous computing, with stronger extensions for multiple technologies, including vector, matrix, and spatial processors. This aimed to promote a richer and harmonized developer community, and a stronger foundation to effectively compete against x86 in the HPC and AI markets. Now, it remains to be seen how Arm will be able to achieve this objective without NVIDIA’s dollars and expertise.

In summary, the failure of NVIDIA to acquire Arm is opening a new chapter in Arm’s future, starting with an imminent, long, and costly process to reach an IPO for the company. The coming years will be probably the most challenging ones in the entire company’s history, leading to uncertainties about its future and its ability to innovate. These challenges may constrain the company from expanding its business to enter new areas, such as HPC and AI. During the IPO transition, Arm will also face serious challenges in its core business, which is the mobile market. Arm is perceived as a foundation among mobile innovators, as the entire industry relies on its ISAs and micro-architectures to sustain technology advancements in many areas, from connectivity to computing technologies. However, any failure to keep up with the innovation pace during the IPO transition could lead to serious fragmentation of the entire mobile ecosystem. So, the looming struggles of Arm is not only about one company, but about the future of an entire ecosystem: the mobile market.