Not since 2006, when CPU company AMD bought GPU company Ati, have we seen such a huge shakeup in the industry for manufacturing computing components as we have today, as Nvidia has announced that they are going to be buying microprocessor designer Arm from Japanese conglomerate Softbank, expanding from being a GPU company into also being a CPU company. The price Nvidia is paying is a whopping $40 billion, made of a mix of hard cash, some to be paid immediately, some following hitting certain performance targets following the acquisition. Usually, when an acquisition is announced like this, it is a pretty hard lock for going ahead, but this purchase will be subject to various different international regulatory checks.
Arm is perhaps not as well known as other giants in the hardware industry, they’re not a consumer-facing brand, and don’t sell any standalone products directly to consumer markets, but there’s a pretty decent chance you own some kind of device using a processor based on their designs.
Arm-based CPUs absolutely dominate the mobile space, with the vast majority of Android phones and tablets, and all Apple iPhones and iPads using chips based on Arm designs. They also power many other devices such as Smart TVs, set top boxes, the Nintendo Switch, Bluetooth headphones, servers, in-car computer systems, and they are even making early inroads into traditional laptop and desktop computers. This multi-industry dominance, and flexibility, is no doubt what attracted Nvidia to this acquisition, and will make Nvidia a serious force to be reckoned with in areas beyond just computer graphics. If I were AMD or Intel, I’d perhaps be worried about what the future could hold.
One of the interesting things about Arm compared to the likes of Intel and AMD, is that they are a microprocessor designer, but not a CPU manufacturer. This subtle distinction is down to the fact that Arm license out their chip designs to many different processor manufacturers, who then in turn design and manufacture processors to suit their specific needs. Manufacturers like Qualcomm, Apple, Samsung, and of course Nvidia have historically taken Arm microprocessor designs and used them to build their own processors.
Perhaps one of the biggest questions is to what extent Nvidia intend to continue this relatively open licensing model, versus hoarding Arm’s technology and expertise to themselves. At least according to the letter of their statement on the acquisition, Nvidia say that this aspect of Arm’s operation will continue uninterrupted:
“As part of NVIDIA, Arm will continue to operate its open-licensing model while maintaining the global customer neutrality that has been foundational to its success, with 180 billion chips shipped to-date by its licensees. Arm partners will also benefit from both companies’ offerings, including NVIDIA’s numerous innovations.”
They also intend to continue to honor existing contracts, use the Arm name, and maintain their UK base of operations:
“SoftBank and Arm are fully committed to satisfying the undertakings made by SoftBank when it acquired Arm in 2016, which are scheduled to complete in September 2021. Following the closing of the transaction, NVIDIA intends to retain the name and strong brand identity of Arm and expand its base in Cambridge. Arm’s intellectual property will remain registered in the U.K.“
With massive acquisitions like there, there are often conflicting drives to use the acquisition to shake things up internally, but also to maintain a sense of continuity in the period of time following the change of ownership. It’s not going to be easy for Nvidia to reconcile these conflicting forces.
For the potentially biggest changes going forward, two areas that we most expect to see Nvidia to use this acquisition to push further with are their relationships to Microsoft and to Apple.
With Microsoft, the potential change going forward could be with regards to Windows 10 on Arm. For now, this is an area that Microsoft has dabbled in, offering a small number of devices from a small number of partners, that run a dedicated version of Windows specific to Arm-based devices. These have key advantages of being smaller, lightweight and more power-efficient, but have some pretty major issues with software compatibility and performance. If Nvidia and Microsoft could work together to resolve some of these substantial issues, there’s no reason that we couldn’t see the first real challenge to the x64 architecture that both Intel and AMD use that has completely dominated the Windows space for the longest time. Whilst the main advantage based on present-day components would be that Arm chips are great for portable computing, there’s no reason that future Arm-based processors instead built to prioritize raw performance couldn’t exist.
For Apple, it’s perhaps more complex. We can help but wonder if the management over at Apple saw this curveball coming or not. They are in the process of pivoting their entire product lineup to Arm-based processors. They’ve had great success in the phone and tablet space with their iPhones and iPads powered by Arm-based processors, and they have already announced plans to end their partnership with Intel for laptops and desktops, in favor of Arm-based chips they will manufacture themselves going forward. Apple has famously very deliberately not worked with Nvidia for their existing lineup of Mac laptops and desktops, only offering AMD based graphics, but perhaps now their hand will be forced. Apple is a company that really doesn’t like their fate to be tied to anyone else, and likely only ever adopted Arm microprocessor designs into its product lineup because Arm has been a very open and platform-agnostic partner. Will this seismic industry shift alter Apple’s plans going forward? Will we get more Intel-based Macs for the time being, whilst Apple reconsiders their approach? Hard to say, but few could have predicted this move from Nvidia well in advance.
Despite these lingering questions related to the makeup of the tech industry as things stands today, the main focus of the announcement from Nvidia is not regarding existing computing hardware, but about building the future, with a big emphasis on AI:
“The combination brings together NVIDIA’s leading AI computing platform with Arm’s vast ecosystem to create the premier computing company for the age of artificial intelligence, accelerating innovation while expanding into large, high-growth markets.”
“NVIDIA will invest in a state-of-the-art, Arm-powered AI supercomputer, training facilities for developers and a startup incubator, which will attract world-class research talent and create a platform for innovation and industry partnerships in fields such as healthcare, robotics and self-driving cars.”
AI supercomputer? Didn’t I see an Arnold Schwarzenegger movie about that once? Regardless of concerns about the growth of AI, whether coming from rational and well-researched arguments, or paranoid delusions of someone who’s seen too many sci-fi movies, there is certainly a business case for investing in this area, and no doubt that’s what Nvidia is betting on here. Gotta make that $40 billion back somehow.
Nvidia is certainly on a roll recently, having wowed the world with their announcement of their latest generation of desktop GPUs, the GeForce 30-series, headlined by the flagship RTX 3090, along with the mid-range RTX 3080, and entry-level RTX 3070. They’ve also updated their impressive DLSS technology to offer AI-based upscaling to resolutions of up to 8k.
No doubt Nvidia has been looking for smart ways to use their huge profits from recent years to expand their growth opportunities, but this is a huge bet, and it’s going to take years before we can see the full ramifications of this move.