For years, Arm, the SoftBank-owned chip design company, was happy to collect crumbs from the table of some of the world’s most powerful semiconductor companies. It levied tiny charges for its designs, bringing it an average of only 9 cents for each of the 30bn new computing devices that depended on its technology last year.
Now, as it returns to life as a public company, Arm has a different idea in mind: It wants to claim a much bigger slice of the semiconductor industry’s cake for itself.
“Every year, more chips ship with Arm and more Arm technology goes inside those chips,” Arm’s chief executive Rene Haas told the Financial Times. “So we’ve got a compounding effect. We think in the years to come we’ll have a good space for growth.”
The pursuit of a bigger share has become central to Arm’s strategy, and to the $65bn valuation the company achieved on its first day of trading on Thursday. It represents more than 20 times Arm’s revenues from last year, at a time when the smartphone market, which accounts for most of its business, has been contracting.
Haas said the Cambridge-based company was now “far more diversified”, having shifted from earning about two-thirds of its revenues from mobile phone chips when SoftBank acquired it to less than half. “We did a lot of work in the years between 2016 and 2023 to transform the company,” he said.
The plan is very different from the one SoftBank’s chief Masayoshi Son had in mind when he agreed to pay $32bn to acquire Arm in 2016. At the time, he dismissed the idea of raising Arm’s prices to justify the hefty premium paid for a company with revenue of only $1.49bn the year before. Instead, the Japanese group’s ambitions were tied to the Internet of Things, where an exponential increase in the number of connected devices, rather than higher prices, was expected to bring growth.
While this remains one of Arm’s four market segments, its attention has swung towards data centre and automotive markets, which are projected to grow much faster and deliver fatter profits. It is also out to claim a higher share of the revenue it helps to generate for smartphone makers.
The new strategy has shone a spotlight on an unresolved issue that predates SoftBank’s ownership: where can Arm find new sales as its core smartphone business matures?
“This has been their challenge for 10 years: ‘How do we get growth in our traditional business and how do we build a second business?’” said one former Arm executive. “They never managed to find their second business, and the base business makes good profits, but they can’t find a way to grow it.”
Arm’s $2.7bn in revenue last year amounted to only 1.3 per cent of the value of all the chips that could benefit from its technology, according to chief financial officer Jason Child.
Taking a bigger share depends on finding new types of customers for its technology, while at the same time persuading all customers to give up a bigger share of sales that depend on Arm technology. Rather than relying solely on traditional chipmakers, it has turned to licensing handset makers, cloud services companies and others, many of which have turned to developing their own proprietary chips.
In one sign of how central to its business this has become, Arm revealed in a regulatory filing last month that Apple has signed a new licence that stretches until 2040. It also claimed a 10 per cent share of the cloud computing market, thanks largely to Amazon, which uses Arm’s designs for the Graviton chips used in its AWS unit.
At the same time, Arm has been trying to extract higher prices. The effort to boost royalty rates followed the introduction two years ago of the latest generation of its technology, Armv9, and echoed previous attempts to use transitions to justify higher prices, said Sara Russo, an analyst at Bernstein.
“The devices that use Arm are becoming incredibly complex,” said Haas. “When you go from 10 nanometre geometries to 5nm and below, the amount of time it takes to design a chip is going up greatly. Additionally, the amount of time it takes to build that chip — to get it through the fab, get it packaged and assembled — is taking longer.”
Arm’s “compute subsystem”, which packages a range of technology and designs for making a complex chip, can accelerate those processes and speed up time to market, he said. “Customers will pay a higher royalty rate for that.”
One banker who advised Arm on the deal said pitching an IPO to investors on a company that develops IP and architectures for chips was “not easy”.
“They’ve done a fantastic job of repositioning themselves,” the person said.
Russo said the company has been working harder to find ways to make its technology more valuable to existing smartphone customers. That has included coming up with two new contracts designed to encourage customers to try out a wider range of Arm’s technology, while also making more regular subscription-based payments.
Taken together, these efforts may already have enabled Arm to raise royalties on new contracts, justifying its confidence in predicting a return to double-digit growth this year, Russo added.
“Arm can charge what they like — and they are,” said the former executive. Pushing for higher prices, however, also increases the danger that some customers will consider abandoning Arm altogether and switching to Risc-V, a rival open-source technology, this person said.
The fact that 10 of Arm’s biggest customers, including Apple and Nvidia, bought more than $700mn of stock between them in the IPO should lessen this concern, according to some observers. The presence of big strategic investors “sends a good message about their future”, said Pat Moorhead, a US chip analyst. “A company like Apple making an investment shows they do think Arm is in their future.”
Additional reporting by Nicholas Megaw in New York