There seldom are two companies so complimentary and ubiquitous in their areas of expertise. But here we are ARM and Nvidia are like twins but differing interests. But this acquisition wouldn’t have probably happened had it not been for the pandemic. SoftBank which was stricken by the fallout of the aborted WeWork IPO and general state of a lot of companies it had invested in thanks to the pandemic had been cash strapped which is why it is selling ARM to Nvidia. Nvidia on the flip side was one of the companies that gained during the pandemic as people played more games, streamed more movies and doubled down on digital. As of today, Nvidia is the most valuable semiconductor company in the world.
ARM is the most important company you don’t know about while Nvidia is the 800-pound gorilla in the graphics space known mostly to gamers. ARM’s technology powers everything that has an “on” switch, be it a toaster or an iPhone and soon even the Mac. Nvidia has been the engine of Google and Facebook’s deep prowess in AI, it has been the engine for online gaming and crazy computer graphics that are now seen in Hollywood movies. One is so loved by Apple that the Cupertino based company has developed all its silicon on ARM’s technology including the new chips for the Mac, while the other is disliked by Apple that it went ahead and used inferior technology from AMD for years and soon will have its own graphics technology. That’s how different and complementary they are.
The differences between them bring forth incredible opportunity as together they could transform the world around us. But also at the same time, this acquisition represents a swath of challenges. Nvidia doesn’t want to alienate ARM’s customers as it is paying a huge sum to SoftBank. At the same time, it wants to infuse its technologies and dive in with ARM in areas where it will end of competing with some of its partners like Intel.
ARM’s ubiquitousness will enable Nvidia to infuse its GPU and AI technology at the heart of ARM’s instruction set and designs. Nvidia failed with its Tegra line of SoCs but in order to incorporate this acquisition, it will likely seek to only infuse its technology with ARM, not create another ARM-based competitor for the likes of Qualcomm, Samsung, MediaTek and Apple. Right now, most ARM licensees use their own graphics stack. Qualcomm has Adreno, Apple has its own proprietary architecture while Samsung is soon shifting to AMD. ARM’s Mali graphics are its weakness, but graphics is what Nvidia is known for.
ARM’s technology will be supercharged in the data centre. Nvidia’s technology has already powered Google and Facebook’s AI efforts in the data centre. In fact, it has become the critical driver for the pervasiveness of machine learning across the technology landscape. But these GPUs were often paired with Intel’s x86 based silicon, now Nvidia can provide a more efficient and affordable offering which will likely be a fusion of ARM and Nvidia technology. Behind the scenes, the underlying computing architecture of the world is moving away from Intel to Nvidia. It has already happened on the phone, now it will happen to compute-intensive things like servers.
While all this happens, ARM’s overtures in the PC space will likely become even more aggressive. Apple’s already made the switch to ARM for macOS and future Macs. Apple’s use of technology-based on Imagination technologies will likely not make it use any graphics solution created by Nvidia but that will regardless supercharge adoption in the PC space considering everything for general-purpose computing is done within the browser.
The biggest thing here is that Nvidia keeps ARM a neutral platform something Jensen Huang has eluded to. If that happens, despite Nvidia’s frosty relationship with Apple, there shouldn’t be any hiccups for both the tech giants. In fact, using Nvidia’s huge scale and cash pile, ARM can get into new areas and can expand its already elaborate customer base which will fuel the licensing model even more.
Nvidia has said that it will be making investments in Cambridge where ARM is based out of. It will remain a UK registered company. It will add another R&D facility to fuel overtures into AI, self-driving cars, healthcare and education. At the core of it, Nvidia and ARM together form the most formidable duo for the future of computing.
This acquisition isn’t without challenges. ARM in China is still controlled by the Chinese. Now that ARM will have American ownership as opposed to Japanese ownership under SoftBank, things could be tricky to get the M&A through in China considering the situation with Huawei and TikTok. Nvidia will also need to manage the situation with Apple quite delicately. Similarly, Nvidia was a big rival of Qualcomm’s between 2011-2014 in the chipset space — it needs to make sure that it doesn’t impact the status quo negatively.
Words : Sahil Mohan Gupta