NokiMo
MooresLawIsDead
MooresLawIsDead

patreon


Corporate Consolidation - Die Shrink Telegrams

When AMD bought ATI, many said either that it was a silly decision by AMD, or that it was a prude decision that would pay-off in a decade...not that many cared about the consolidation aspect.

Yet here we are in 2021 - Nvidia is trying to buy ARM, and Intel is gobbling up tons of smaller companies and groups of engineers.  Somewhat related, Microsoft is buying up tons of developers as well to seemingly force their service into a success through monopoly.   What do we think about it?


That's the subject of the next Die Shrink - What do you all think about Nvidia Buying ARM?  AMD-ATI having a GPU+CPU synergistic advantage?  Intel and Apple buying up engineers?

When is it good to combine and grow?  When is it bad for the industry?


You have roughly 14 hours to submit questions!

Comments

With the upcoming Xe architecture of Intel I hope we will see some decent competition on the APU market since the dedicated gpu situation probably won't improve for quite some time. As a conclusion, I can say that for AMD the purchase of ATI was a great investment. It might has even safed AMD during the dark Bulldozer times since they had still some decent business opportunities thought their business on APUs for consoles. I'm still a bit disappointed why AMD hasn't updated their APU with a combination of Zen3 and some RDNA2 CU yet. The reason might just be that AMD is still benefiting too much by the insane demand for dedicated GPUs.

Kai-Uwe Markeli

I wanted to give my perspective as someone whose company was recently acquired by a Fortune 20 company. I think Nvidia/ARM is a different situation than the smaller acquisitions to get smaller IP or dev teams. A lot of companies in the tech space are looking to get acquired and talented people are generally always looking to "prove" themselves at a large company. While we do see some downsizing after acquisitions, my whole team was kept by the acquiring firm and all of us have done well after acquisition. I look at those type of deals as generally a win/win on both sides (even our admin people who werent kept got a pretty nice severence package). ARM is a different beast as I think it's more of an anti-competitive thing, Nvidia is clearly looking to consolidate to become a competitor to AMD/Intel which makes sense strategically but doesn't seem like it will be good for current ARM customers or the market (The only way this deal makes sense is for them to try to do the acquisition to drive up pricing, otherwise they would just be a customer).

B. Fish

Considering how much Microsoft has previously been 'attacked' for anti-competitively shipping their own default browser in their own operating system (something I personally believe they deserve to do given the effort of building an entire OS), do you see any unfairness or hypocracy that other operating systems with their default browsers seem to raise no eyebrows (i.e. Safari on Mac OS and IOS, Chrome on Android)? While MS Edge may be a reasonably capable browser these days, the seeming desperation of Microsoft to get more people using it is the only reason why I don't want to use it. What does MS seek to gain from me using Edge? What data must it be collecting (and couldn't they use their OS to collect the same data anyway)?

Ryzza5

I think AMD buying Xilinx is a really important acquisition for the industry. With that deal through It will be a very formidable foe for Intel and Nvidia in 2024-25 in terms of size and technology and that's important for the customers.

QuickJumper

In the SemiWiki podcast #28, Daniel Nenni and his guest (Wally Rhines) discuss business strategies for startups in the semiconductor space, one of the points they talk about is what they call the exit strategy. Here, Rhines notes that the main two options for startups are either to make an IPO and become their own independent company, or to look to be acquired by a larger player in the industry so they can use the infrastructure of those larger companies to get their products on the market. Although I agree that many companies who have a long history in the industry have been acquired over the last few decades, which is a substantial degree of consolidation, a lot of the day to day news/drama about monopolism and excessive consolidation tends to mention those startups. Do you think that lumping in aqcuisitions of those startups contributes to an excess of "hype" around acquisitions? Also, I feel like whole the last 2 decades have seen a lot of consolidation and optimisation in the semiconductor and computing space, I feel like the industry is changing gears again and being more creative with solutions, which will drive up diversity and make space for new players in the industry. I believe we are already seeing this with the HPC accelerator and server markets.

qhfreddy

I think the Nvidia+ ARM acquisition makes sense in theory. If ARM wasn’t its own monopoly I really do think it makes sense for Nvidia to have them under their wing. But we all know Nvidia needs to control everything they touch and it would creep into how ARM is run even if they say it wouldn’t. I think in Microsoft’s case I am completely fine with them buying up studios and developers as they are using it to bulk up their gamepass service. Gamepass is a really good value for gamers on both xbox and pc. My position would be if these acquisitions create tangible benefits for me as a consumer, I am fine with it.

j Steez

I think further consolidation is inevitable and even desirable up to a point. If a corporation buys a smaller company, -directly related to their core business- it will usually benefit. At the very least, they remove a competitor and gain market share. If they gain some useful IP and staff, it's even better. The consumer may benefit, as huge corporations can afford to spend more on R&D, knowing they will ship high enough volumes for this to pay off. This may even be necessary given it is getting more and more expensive to improve technology. However, buying an unrelated company would be less useful. For example if a CPU maker bought an anti virus company, that would be really stupid.

Laz

If monopolies stifling innovation became a real concern, do you think governments would step in? Or are these tech giants too powerful to be affected by anti trust regulations?

Laz

Can you think of other industries, which became heavily consolidated in the past? How did it all work out?

Laz

Let's also not forget about Mellanox Technologies, to date, Nvidia's largest acquisition. This absolutely strengthened their HPC segment, providing high-speed network and interconnect technologies for nvidia. I believe they are using their technology for A100 interconnects within a chassis (think Infinity fabric), but most certainly the communication between nodes... but this only indirectly affects consumers, for now, as these are datacenter technologies. But as for Arm, I fear that this is mainly a negative thing for consumers. For nvidia it rounds out their offerings in the datacenter offerings, but the closed source nature of nvidia is only going to have negative impact on such a large (more or less) open source community. If nvidia didn't have a history of leading the way to drive prices up, it would not be as big of an issue. So what are the effects of Arm likely being owned by nvidia? - More expensive phones. - Crippling low-cost development of aarch64, like the raspberry pi and other SBC's. - Goodbye to the low cost Wyze products many have come to enjoy. - And that is just the beginning, Televisions, streaming devices, network devices, car infotainment... And the list goes on. What are your thoughts on the impact of this wide grip across so many consumer products when nvidia hasn't had the best track record in trying to drive costs down?

Well there you have your own answer, it's only because Sony is keeping Microsoft on its toes! 😉

Eleriam

On the topic of corporate consolidation, I've three discussion points. Nr. 1 With the imminent acquisition of Xilinx by Amd for FPGAs and their use in the next architectures of Zen in what sector do you think, will be the next big buy of AMD for a tighter intergration/advantage? Where to you think is Lisa Su looking next? Nr. 2 Do you think that AMD could become a target for a acquisition by a bigger fish. For example an entrance into the b2b/b2c market for a foundry (TSMC/Samsung). With a acquisition of AMD by TSMC the fight with Intel would be on, as they would be both having the same portfolio/business. Nr. 3 Global Foundries had their IPO almost a month ago in a quiet fashion. Where do you think is their place in this accelerating market? As Amd is phasing out their products there, can they survive on the older notes with low margin chips? Have you heard anything exciting from them?

Polarde

It is very concerning. I always say “vote with your wallet”, but if there is no alternative then that is no longer an option, or perhaps you choose the lesser of two evils. However, that’s what everything is turning into right? An illusion of choice.

DeadOfKnight

What do you think of the Activision-Blizzard being Merten in retrospect? Way back when I thought Activision would tarnish Blizzard's name, but since the controversies this year, I can't help but think that maybe Blizzard is tarnishing Activision's name?

I'm not a big fan of the acquisition hungry nature of the tech industry, on both the hardware and software side. Besides the fact that less competition means less drive to improve their products/services, it also means there's less competition for fabs and users, which drives prices up in the long run. That said, the main cause of this is decoupling stock valuation from anything that is actually verifiably real, so I'm not surprised that big tech companies feel its cheaper and easier to gobble up smaller companies to boost their valuation over just getting into license agreements to use their IP. Being able to buy other companies is now a sign that your company is strong, because it means you have money.

Cleansweep

Hey, here's a controversial hot take: presuming no monopoly, defined in the Sherman Antitrust act based off of percentage of the market owned (i.e. Intel), consolidation comes with some great benefits that people pretend to ignore. For instance, if any other company owned Halo, we all know they would have shoved infinite out the door cyberpunk style. But they didn't and now we're all getting a game that I would argue needed Microsoft to exist as is. And they'll have to keep doing that, because they don't own the majority of the gaming market. After watching my favorite small dev (Obsidian) get shit on for its entire existence, and seeing AAA gaming wither and die under the stewardship of the big publishers, I hold no romanticism in independent studios. At the end of the day you have to deliver, and Microsoft has a better motivation to deliver than Activision, so long as Sony keeps them on their toes.

Tick Dickler

I know nvidia are trying very hard to buy arm (i havent kept up with all the details) but the a100 shows they have no trouble licensing arm chips like everyone else (I know they're not apple) but could nvidia not simply branch out into creating logic chips (seems there should be some cross over between gpu and cpu chips) That way they have their own proprietary hardware no monopoly concerns, may take them a little longer but they must have the money...

Alex

Hey Tom and Dan the man with all the plans, I personally like the AMD and ATI merger it makes sense, as does the ARM and Nvidia deal. Having the ability to create solid hardware SOCs as opposed to discrete chips makes sense. I mean Yes Intel Integrated Graphics sucks to game in but for someone doing simple doc creation and looking at the BookFace it’s fine. So here is my question to ponder, if Intel Integrated Graphics become as good as say Nvidia MX graphics in the next few years, do you think Nvidia(if the arm deal goes through) would be able to make a competitive product or would Nvidia be able to make a better product if they just focused on GPU technology?

Brad Medlin

Hey Tom and Dan. I do not have much to add as I am terrible at thinking of questions. However I wonder what y'all have to say about the ARM acquisition. Do you see Nvidia continuing to license it out to other companies or do you see them repeating history to Gain an edge over the CPU market?

Nvidia currently operates on a aggressive proprietary manner which makes it incompatible within the Android ecosystem. Android, is very varied, has enabled it to bite effectively on heels of X86. I feel Nvidia in control of Arm would restrict Arm's way of working to ensure the gap between X86 and Arm will be restricted if it effects Nvidia's X86 sales. Arm's market success is down to it open collaborative nature which is the antivisis of how Nvidia operates.

Baza's Sami Note II

I don't have much to contribute on Nvidia/Arm other than, good gods, please anything but Nvidia buying Arm. Please keep the iron grip of Nvidia away from the only mass-market neutral party we have.

Gwen Farron

Burning the midnight oil Tom? I understand the cons to Nvidia acquiring ARM. What would the pros to this acquisition be? Would we eventually see a competitor to AMD, Intel, and Apple in the consumer CPU market? Would this lead to truly great APUs that would allow us to ditch bulky graphics cards?


Related Creators