Opinion: Will someone else buy Arm?

Jay Goldberg

Posts: 52   +1
Something to look forward to: There are rumors that Qualcomm, among others, are looking to buy Arm. After the failed saga of Nvidia trying to buy Arm, it looks largely impossible that any other company could run the regulatory gauntlet. For its part, Arm does not seem terribly enthusiastic about going public later this year or next, but seems to have no other options. That being said, there is one scenario that could square this circle -- a consortium of major Arm licensees could buy Arm. This could potentially include Qualcomm, Mediatek, Nvidia, Broadcom, AMD, Tsinghua Unisoc, among others.

In theory, this would address everyone's concerns. Softbank could get an exit from its investment, something that it has repeatedly signaled it is eager to obtain. Arm's executives could remain free from the scrutiny of public markets, and theoretically secure funding for its strategic plans, whatever those may be. And Arm licensees could get assurance that Arm progresses on its much-needed roadmap, without having a competitor dominate that roadmap to everyone else's detriment.

Editor's Note:
Guest author Jonathan Goldberg is the founder of D2D Advisory, a multi-functional consulting firm. Jonathan has developed growth strategies and alliances for companies in the mobile, networking, gaming, and software industries.

The practicalities, however, are far less straightforward. The first concern would be under what structure would the company exist. Softbank wants an exit, and no one else would want to consolidate Arm in its own financials, so the company would have to become some standalone, private entity. This likely involves inviting in a private equity investor. This would bring with it the added benefit of having a 'neutral' party as the leading voice on the Board. But any such investor will primarily be motivated by financial returns, as opposed to the strategic needs of everyone else in the consortium.

As with any scenario, it is unclear how or if Arm would acquire the capital increase it has repeatedly said it needs, private equity is not known for R&D investment. So ultimately, Arm is going to have to go public at some point, but maybe this path gives them a much longer time to prepare for that.

A second problem can best be described as 'how to herd a bunch of cats'. That list of potential participants is very diverse, with a wide range of competing interests. Each of them would participate on the principle of protecting their individual strategic interest, and thus presumably want to have some influence on the company. This runs counter to the idea that the new company is independent.

This is not an impossible barrier, but it means that aligning everyone's incentives will require some very complex negotiations. Ultimately, the devil will be in the details here – how do the various blocks align and who ends up with a decisive vote that translates into outsize influence down the road.

Another important question is what will it mean for everyone else. We have written extensively about the conflict inherent in Arm's current pricing regime which penalizes new entrants to the benefit of large, established customers. If those large customers end up in control of Arm, then the one thing they can all agree on will be further entrenching that conflict. This matters a lot for the company's long-term prospects, in whose interest will the new owners vote?

And finally, there is the question of China. Our best guess is that China's government will be willing to sign off on the deal if Chinese companies get to participate in the consortium. We added Tsinghua Unisoc to our list above, as they are the most likely candidate, but there are others.

Could Huawei's HiSilicon join? They were once one of the top Arm licensees, and show signs of re-emerging from under the US Government's proscriptions against them. Would they be allowed to participate? And since the resolution of the Arm China soap opera, Arm has been on a tear there signing a lot of deals, how will all those companies feel about some new group of American owners?

Ultimately, we suspect the recurrence of these rumors means that someone, somewhere is negotiating this kind of deal. Arm's bankers have to at least entertain the idea as an alternative to an IPO. So of course, the question will likely come down to valuation in the end.

The public markets have already given some pretty clear signals as to the value of an Arm IPO, and it is well below the $40-ish billion Nvidia offered. Would the consortium be willing to pay above the Street? The answer to that will depend on how much of a premium they will pay for control of the company, which of course goes against everything we wrote above. If we had to guess, we would say this is the company's preferred option, but it is complex enough to mean this outcome is far from certain.

Permalink to story.



Posts: 4,416   +7,988
Qualcomm buying ARM would pose the same issue as nvidia - a licensee now owns the IP and can deny it to others. Why does ARM need bought anyway? They have money coming in from every ARM makers on the planet. Smells like the C suite wants a retirement check.


Posts: 747   +351
ARM became what it became because it sell licenses so others can build on it's IP... A one side controlled ARM I think at long term would make happen the same that Apple did to Intel, as it switched Intel cpus for in-house developed arm processors... A harder and costier ARM solution would make alternatives more attractive.

In the end it's a balancing act.

Mr Majestyk

Posts: 1,729   +1,600
Agreed, Qualcomm and Nvidia should not be part of any consortium. We do not want those that rely on Arm's IP for so much of their product stack to be involved. They need to be at ARM's length. We need firms that do not have vested interests in controlling ARM for their own good. A duopoly is not really any better than a monopoly.


Posts: 852   +745
You didn't understand: ALL those (not just one or two, but all) would form a consortium where ALL of them would rule, so no-one would be left out.
This make little sense to me. If all in the consortium have equal rights, then what is the point of owning part of ARM? I feel ARM is better off by itself, or with a company that will result in no conflict of interest to the IP holders. To me, this is a very sly move to try and acquire a decent % of ARM shares. You cannot outright buy the company since Nvidia have tested and it failed bitterly. So Qualcomm’s tactic seems to be to start small, and who knows over time you start buying ARM shares over from competitors that need liquidity.
Qualcomm buying ARM would pose the same issue as nvidia - a licensee now owns the IP and can deny it to others. Why does ARM need bought anyway? They have money coming in from every ARM makers on the planet. Smells like the C suite wants a retirement check.

ARM's revenues are very low. They only made $2.7 billion in 2021. That was A. a 35% increase over 2021 and B. a record. That is revenue, not profit. Do you actually think that ARM's R&D can keep up with the competition, massive companies like Apple, Intel, AMD, Microsoft, Amazon, Google etc. with revenue that small?


Posts: 456   +363
This make little sense to me. If all in the consortium have equal rights, then what is the point of owning part of ARM? I feel ARM is better off by itself
It makes sense because the owner company wants to sell to make more money for something else. So if one company is not allowed to own ARM, if all present licensees that have more to win or lose want to buy it, it should be easy.

Advantages: ARM is improving designs very little in the last couple of years, Apple does a MUCH better job and is using just ARM compatibility, not the designs. ARM just does designs for a lot longer and is not as effective. Apple is much younger in that and does much better. If Qualcomm, Nvidia, etc put some hand on it, they may stimulate the designs.
For everyone who is singing the "keep ARM independent!" song and dance, ARM gets less than $3 billion a year in revenue. That is REVENUE, not profit. Imagine what the big tech companies - Apple, Intel, AMD, Nvidia, Qualcomm, Microsoft, Google, Samsung, Amazon - spend on R&D.

The only thing protecting ARM Holdings is licensing. You can only design your own ARM cores and CPUs if you have an architectural license. Only a few companies have architectural licenses and the others - including Apple - do not license cores to third parties to make CPUs. The truth is that several companies who own ARM IP - Fujitsu, Ampere and of course Apple - have designed ARM CPU and GPU cores that are much better than ARM's Cortex, Neoverse and Mali. They just don't license it to anyone else. And ARM Holdings doesn't have the R&D budget to compete.

Now RISC-V is open source. Which means that any company can pour billions into making RISC-V based CPUs. This hasn't happened yet, but RISC-V is already eating into ARM's licensing at the low end with things like microcontrollers and IoT SOCs. If 10 - or even 5 - years from now RISC-V is a viable competitor with ARM, what will that do for ARM's licensing revenue and valuation then? And where will all of you who keep demanding that ARM stay independent be?


Posts: 456   +363
. And ARM Holdings doesn't have the R&D budget to compete.
Not everything is R&D budget, the more the better. The persons behind have to have fresh new ideas and not "we're the big players so no messing around.. )