The fight for Arm

The gunfighters are circling the corral and their fingers are twitching. In-play is the opportunity presented by Arm.

Kathleen Maher

The gunfighters are circling the corral and their fingers are twitching. In-play is the opportunity presented by Arm. As if it’s not enough that Arm owns mobile phones, automotive, consumer electronics, industrial appliances, and it has a growing role in the datacenter, Apple has proven that Arm can take over in personal computing, and even workstations. There are a lot of people who don’t like that idea.

Nvidia has been poised to take Arm to the promised land since Jensen Huang announced plans to buy Arm, the IP provider, from Softbank in 2020. Since then, an alarmed industry has thrown up nothing but roadblocks and complaints. China, the EU, and Great Britain have all been maneuvering to block the deal.

Meanwhile, Intel has been pulling on the cord over and over again to restart its engine and continue the path it has traditionally taken to mow down the competition. Ironically, in 1999, Intel had a shot at building a complementary business around Arm for mobile and appliances through StrongARM but it lacked the will to compete with itself and its flagship x86. To be fair, Intel didn’t do all that well on its more recent efforts to build a parallel business with AI/ML processors.

Meanwhile, China has been circling a nervous Softbank and asserting its demand to a Chinese version of Arm. Why China thinks it has a right to demand such privilege to privately held and paid for IP is beyond our understanding.

And then there’s Apple. There is a lot to be said about Apple’s strategy of plotting ambushes in public. The company had been promising to build its own chip. It has been amassing the technology for well over a decade but does its hard work in the dark. The competition, with every reason in the world to fear Apple’s ability to transform markets just by entering them, prefers to console itself thinking that Apple is just going to build a bigger iPad, or that all these big plans will end up like the Apple car, or an Apple-branded TV. And then boom! The company does just exactly what it said it was going to do with the M1. Now people are nervously looking around for that car to show up any minute. (Though probably, what they need to be looking for is Apple’s take on the realities, A, V, and X. Apple is slouching toward that direction.)

Apple has clearly demonstrated it can remake the PC industry.

Qualcomm, the mobile powerhouse, the king of the Android phone, announced at its recent investor day event, its intention to counter Apple with a powerful CPU. Apple is not the only company that knows how to build in the dark and Qualcomm has plenty of raw material to work with including its own Arm-based processor, Kyro. Qualcomm has gained the support of Microsoft for its alternative to Apple’s M1, while Intel has just put a metaphorical head in metaphorical hands and sighed—another one, great.

So, let’s run the movie forward. In one scenario, there are bodies on the ground. In the other, everybody wins.

What if Nvidia succeeds in its pursuit of Arm? It will be bound to maintain Arm’s availability and neutrality. Qualcomm and Microsoft have the potential to build a powerful PC business. Intel continues along its way but could well start to decline as it’s hemmed in by its competitors. And Apple drives away with an invigorated new market across a dramatically changed landscape.

Arm is ironically a victim of its own success. It has gotten so integrated and ubiquitous it has become an international standard in which everyone thinks they have a stake. Nvidia is a powerful and clever company, and not one to shrink from a fight—but it’s tough to take on everyone at once. It’s not only tough, but it would also be foolish. Nvidia has agreed to pay SoftBank Group $1.25 billion if the planned sale of the chip design house Arm Holdings to Nvidia fails to close. The question Nvidia has to figure out is: will it cost more than $1.25 billion to pursue it? And how do you measure the distraction of management’s time giving depositions and sitting in meetings with lawyers and shareholders? What won’t be getting done because of those distractions? AMD and Intel probably want Nvidia to single-mindedly pursue Arm. That’ll keep Nvidia busy and take some of the pressure off them.