24-11-2021 | | By Robin Mitchell
The long-awaited Nvidia acquisition of ARM has again faced a new delay in the UK, with investigators looking into national security concerns. Why has there been significant amounts of pushback against the acquisition, what concerns investigators now have, and is the acquisition a good idea?
The acquisition of ARM by Nvidia has been in the works for over a year, and the numerous setbacks and delays witnessed have many wondering whether or not the agreement will ever be completed. The worldwide nature of ARM makes it difficult for Nvidia to compete with various countries worldwide, including the United States, the United Kingdom, European Union, and China. The current semiconductor shortages have made many involved more worried than ever when it comes to national security concerns.
When the deal was announced, many worldwide, both individuals and companies, publicly announced their disapproval of the acquisition. This large amount of negativity has many factors, but the two most prominent relate to the neutrality of ARM and the risk of US law being imposed on customers worldwide.
ARM provides companies with licenses to implement and manufacture ARM core processors into their designs while simultaneously providing software resources for targeting their architecture (i.e., C++ compilers etc.). One of ARMs core principles is that no customer is prioritised over any other customer. Anyone wanting to use an ARM core in their design can do so with no discrimination.
ARM can be neutral with its customers as ARM is not owned by anyone tech firm whose products and services have a technical interest in ARM products (i.e., little to no conflict of interest between the decision-making process and the use of their products). However, Nvidia taking over ARM would see this change immediately as Nvidia themselves are customers of ARM. If Nvidia was to have control over the technical side of the business, it could potentially direct ARM developments to better work with Nvidia products such as their GPUs. Furthermore, Nvidia could also prevent ARM from selling to its competitors or give itself early access to technological developments in ARM.
It is also important to note that ARM is based in the UK, meaning that ARM operates under UK law. Nvidia, however, is a US-based company meaning that if Nvidia was to acquire ARM then suddenly US laws could apply to ARM. If this was to happen, customers from countries such as Russia and China could find themselves unable to acquire important ARM technology that they currently rely on as the US has imposed multiple trade bans on advanced technology.
Just when things couldn’t get worse for the Nvidia ARM acquisition, the deal is now being invested in the UK on national security grounds. The UK’s secretary of state for digital, culture, and media, Nadine Dorries, has instructed an investigation into the Nvidia takeover of ARM after a report on the deal was released in July by the Competition and Markets Authority.
While little information has been shared as to the exact nature of concerns, it is most likely to do with the advances that ARM makes in the field of microprocessor technology. ARM cores are based on RISC architecture, but despite their simplified instruction sets, they can still perform heavy tasks efficiently. In fact, ARM cores are now being considered in use with servers and computing clusters such as the Fujitsu A64FX supercomputer, which integrates over 150,000 ARM cores.
The recent shortage of semiconductors around the entire planet has demonstrated how dependent nations are on semiconductors. A nation that cannot access semiconductors cannot produce tech, and a nation that cannot produce tech cannot compete on any front with any tech-enabled nation.
Thus, the UK having control of a company that can produce advanced processor architectures capable of running the most demanding tasks means that the UK will never be dependent on an outside nation for technology (specifically, processors). If the unthinkable happened where nations worldwide closed their trading borders, the UK at least has the means to design processor technologies that are critical for defence technology. Giving control of ARM to the US under Nvidia not only sees the US have direct access and control over ARM but could even allow ARM to be sold again to other countries such as Russia and China.
Many semiconductor companies buy other semiconductor companies, and these acquisitions almost always involve the two combining their resources to produce new products. In most cases, this can be highly advantageous as combining technologies can produce even better products (such as a Silicon Carbide company joining forces with an instrumentation company to produce power op-amps).
However, Nvidia acquiring ARM would only lead to better products if ARM and Nvidia join forces on GPU and CPU technology. This joining would automatically see Nvidia preferred over its competition (such as AMD and Intel), thereby voiding Nvidia’s promise to keep ARM neutral.
The only way that Nvidia can remain neutral is to not allow the two companies to share information or combine resources, which begs the question, “what is the point in purchasing ARM”. If the acquisition is to simply acquire wealth, there are plenty of other semiconductor companies who would present Nvidia with a conflict-of-interest-free acquisition.
Overall, Nvidia acquiring ARM seems like a terrible idea that will disrupt the entire ARM supply chain. Of course, this acquisition could have been a simple ruse to try and get coverage on both companies as Nvidia faces challenges from AMD and Intel while ARM faces challenges from RISC-V.