FTC Suing NVIDIA Over Arm Deal
by Daniel Newman | December 14, 2021

The Six Five team discusses the FTC suing NVIDIA over the Arm deal.

Watch the clip here:

If you are interested in watching the full episode you can check it out here.

Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we do not ask that you treat us as such.


Patrick Moorhead: Let’s jump into FTC suing NVIDIA over the Arm deal. So it’s interesting. Just to put some of this in perspective, a couple years ago, maybe it was 18 months ago, NVIDIA made a play for Arm and they have been slowly making their way across the globe to get approval for the deal.

Arm is essentially the arms dealer for the smartphone market, IoT market, and increasingly becoming relevant in areas like ADAS, even the server market. Gosh, if you look at what Amazon is doing with arm-based processors and the PC market. So to stop the deal from going through in the US, the FTC sues. It sounds pretty dramatic. Okay?

But it’s essentially this is the FTCs instrument to slow down this deal. And they were very focused on areas that they would harm competition, which were ADAS, smart NICs, and Arm-base CPUs. So you can imagine who might be complaining about this? And you have to have complainants in this.

And ultimately we will find who it is, but I wouldn’t doubt if it was Intel, Marvel, who’s in smart NICs, multiple small ADAS systems, maybe even Mobileye, and potentially Amazon with Arm-based CPU for cloud service providers. So I’m not naming names here, but this is likely who’s behind this.

The irony is always the FTC comes in and talks about how it harms competition, but nobody was talking about how this potential deal could actually enhance competition. There is no doubt in my mind that this deal will increase competition in the hyper scale server market and quite frankly the enterprise server market, which is dominated by Intel and has been dominated for the last 10 years, with AMD just recently getting into the double digits.

But if you look at Power, if you look at Arm, if you look at a host of other processor companies and instruction sets, they really don’t have any market share at all. And I think probably the second thing NVIDIA would do is just ramp up competition in this market under the NVIDIA banner.

The other place where I think this D yield could increase competition is the PC market. AMD and Intel have a hundred percent of the PC market. Well, let’s say 99.97%. Qualcomm is coming in strong with a new roadmap. When the NUVIA architecture hits, they’re going to get more competitive.

But the first thing I think NVIDIA would do if this goes through is to create a monster CPU chip. It’s a market that NVIDIA understands. It has power. It has respect. And I think that is an area where they would actually increase competition. And in the end, really what all these companies want who are complaining is they want a fair shot that NVIDIA won’t get first dibs on the IP.

And secondly, they want certain features in new instruction sets that benefits their business as much as it might benefit NVIDIA SOC business. My final question is, well, what next if this doesn’t go through? We’ve talked a whole lot about I just pontificated on what I thought was going to happen if it moved forward.

What else is going to happen? Is Arm going to go through an IPO and then essentially be starved for resources and have to get mega profitable as it’s been investing for the past five years? Is there going to be some magic consortium of companies who are going to throw down money to make it happen?

Boy, you think getting Arm to make agreements on a new instructions set are hard. Imagine having to get Apple, Samsung, AWS, Qualcomm, Microsoft, and Huawei, all to agree on something, pretty much all competitors. That would be slower than an IEEE group itself. And that’s super slow.

I did hear a very provocative thing, Daniel, that I wanted to share with you. We’re not rumor mongers, but I had dinner and somebody shared something that very rarely makes me do one of these. And that was, “Hey, Patrick, it’ll be interesting if the Arm doesn’t go through.”

QTL from Qualcomm gets spun out and gets combined with Arm and you have a super duper licensing company that goes all the way from the very smallest core to 5G and wifi and connectivity and ends up to be the IP arms dealer of the planet. Anyways, I’ll leave it there. I’m going to pull the pin on the grenade and throw it in room and then leave.

Daniel Newman: I’m not even going to quite comment on that. I can’t imagine how the regulators would look at that setup. I mean, obviously Qualcomm has been through years of its business model being scrutinized, but the recent FTC ruling did rule in its favor and it is going to allow them to continue operating.

Which of course, then you have to put that in contrast to what’s going on here, because that does set up a fairly interesting and provocative discussion, Pat. Here’s the kind of where I’m struggling. So one is Arm doesn’t make a lot of money in its current state.

I think it nets about a billion dollars. And in this business, that’s not very much. It is the current instruction set provider to a number of the world’s largest technology companies that have spent lots of money to in-house the talent to take an Arm instruction set and make it usable at scale, whether that’s AWS and Annapurna, whether that’s what’s going on with Amp here, partnering with companies like Microsoft and Oracle. You’ve got Apple, who of course has done really well with an Arm instruction set on their new max.

Of course, in the beginning, we found all the things they did wrong, but very quickly rectified most of them. There’s still some work to be done. But my point is that it’s a very interesting crossroads. Because the companies that are in the current state of Arm, doing well with Arm are companies that have vastly more resources.

So it’s not really stirring up competition. It’s just enabling the biggest companies to basically take advantage of what Arm has built and they make it better. The promise that NVIDIA’s making is to inject significant R&D capital in to ideally make the instruction sets more off-the-shelf ready. That would be the ideal here.

The problem is, does anybody really trust them to do that and not have some sort of self-interest? Which then, again, takes you back to what I mentioned when I talked about the Qualcomm ruling. Because of course, they’ve got a similar ecosystem that’s got to be looked upon here.

You’ve got an Intel and you’ve got the AMD that’s basically holding the entire x86 market under their control. And there’s no plan to distribute that in any capacity. And at least having a more competitive offering would at least add a third. And Pat, you and I always talk about having three is always better.

So if we’re consistent in our verbiage, then this could potentially help Arm rise up to be a genuine x86 competitor. Whereas right now NVIDIA really doesn’t have any market share to compete with x86. It’s inconsequential the amount of Arm market share.

Even as AWS is doing really well, even as Apple is doing really well, it’s still pretty small in both those categories. Okay. So quick pivot here. This deserves a pivot. The other challenge I have is we’re in this state right now where big tech, you have to ask a question about the regulatory environment with Lina Khan and what she’s doing.

There’s step A, which is appropriate regulation and enforcement, which is something I think we’ve been pretty weak on and we’ve let deals go through. That’s how the Instagram, WhatsApp, Facebook thing has become what it’s become. Now, at the time when Facebook acquired those things, there was no breach of antitrust law and there was no clear indicator of how they would actually breach antitrust law.

But of course, if you’re a conspiracy theorist or you’re just negative on the intention of big tech, you can look at any acquisition deal. This could be AMD Xilinx. And you could say they’re going to control the entire FPGA market at large. We can’t let this deal go through.

Is that really necessary to stop? Is that even a big enough market to warrant stopping? But just to have that power and wheel that kind of control here. So what I worry about here is now we’re going from a state where the FTC needs to look at a deal and see legitimate opportunity to create harm and then stop the deal before it gets done because they know that harm will be done.

But on the other side, it’s their job to review the plan and then hold the company accountable to the plan. I’ve even heard Jensen and NVIDIA has gone as far as to say they’ll create a licensing business that’s a wholly separate company from NVIDIA. So they’ll have the R&D and such in-house and there’ll be a relationship, a little bit like what we talked about with Intel and Mobileye, where they’re going to cross-relationship for certain resources, but it wouldn’t even be held underneath NVIDIA.

And that would give the opportunity to more directly regulate the business. And I heard that Lina khan rejected that. And of course there’s reasons to reject it. This Arm space between mobile connectivity, servers of the future. I mean, this is a massive opportunity. But the idea is now we’re kind of playing the minority report game. We’re basically assuming the worst. We’re looking into the future.

You’re guilty before you’ve broken the law. And we have people who I don’t think are necessarily competent to read the tea leaves. Because if they were, they wouldn’t be working for the government. They’d be building awesome freaking companies. Sorry if you’re the one.

But my point overall is that I worry about how much control lawmakers, legislators, policy makers and influential in the beltway people have to stop a deal, to stop another company from rising up and becoming even stronger, which NVIDIA has done. But at the same time, Pat, I can easily argue either side of this deal.

I can easily argue why they should let it happen and I can easily argue why they shouldn’t. And that’s why for the last six months I’ve called it a coin flip. Now, with that, I will say this deal is probably less than 10% now. The probability of this happening with all of those probes plus the FTC is less than 10%.

But I don’t know what SoftBank’s going to do. And I don’t know that people realize that this deal not coming through, a consortium’s not going to work, maybe an IPO on its own. That would need to raise a ton of capital because the company in its current state isn’t worth that much.

So it’s not going to go raise a ton of money on an IPO unless people really are betting on the future. But where does that R&D money come from, Pat? It’s going to need to make it more competitive than it is today? So SoftBank’s on the hook, NVIDIA could be out billions of dollars. This can weaken the competition, which is the unexpected consequence is, who’s going to pick that up? So it’ll be very interesting to see.

Patrick Moorhead: My favorite one of your comments was the precognition. And if you look at what the FTC did with Qualcomm, they never actually showed any proof of consumer harm. What they showed proof of was potential consumer harm, which was just prick and ironic. Which by the way, back in the day when I was a chief witness on AMD versus Intel, and also all the investigation of Intel, there were mountain of proof and we would’ve been laughed out had we had no proof. It’s crazy.

Daniel Newman: You know what, Pat? Long and short, it’s very interesting because it used to be consumer harm and competition were almost uniquely interdependent. And it’s not the case anymore. This is why platforms can’t be regulated the same way. And you and I have talked about this.

We did an awesome pod back in the day. I’m calling about this in case we can put it in the link. But you can’t actually say that people care. They love the apple platform. They love the Facebook platform. Nobody cares about side loading. That’s only a competitive issue. So now you’ve got this conflict when the company didn’t have enough competition, the consumer was hard.

But now the consumer’s like, “I love this. I don’t care. I don’t want another way to buy this.” I just want to buy it here on the platform. This is the same thing here is it’s not always together. So we might need the spur competition, but you can’t say that consumers are going to be harmed by having more choice and having better product.

So I don’t know. This is going to be a really interesting one to watch, but my gut says it’s over, as much as I hate to say that. I just don’t see how this goes.

Patrick Moorhead: Yeah. I’m not a very good odds maker, but it’s dramatically lower in my mind than where I was before the FTC had. I thought, for sure, this thing would get hung up in China. I figured they could pay enough money and follow the rules in Europe to squeak by. But this might be different. But hey, let’s move to our final topic. I got to tell you, I think I like Saturday pods better, Daniel.

Daniel Newman: It’s fun. It’s fun to have a little more time to boomerang. We’re going to finish probably faster than normally, even though we’ve done more boomeranging. Maybe when we don’t have as many like signals, because on Friday morning, all the emails are flowing and you got lots of things and meetings ahead, where maybe we talk slower. I don’t know. But we’re flying through.

About the Author

Daniel Newman is the Principal Analyst of Futurum Research and the CEO of Broadsuite Media Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise. Read Full Bio