Nvidia has approached Softbank and is considering buying ARM Holdings

[A///]

Hopefully the market shoots their price up.

 

[TheGiant]

@StinkyPinky

To build on what @Stuka87 has said:

Apple, Samsung, and any other design firm that has an unlimited license to ARMv8 will be able to continue making their own ARMv8 CPUs. Samsung has previously halted their in-house Exynos development (in favor of reference ARM designs), but there is no reason why they can't go back.

Any firm that licenses specific designs from ARM will continue to be able to use those designs. A76-A78, etc.

Any future ISA improvements nVidia makes (ARMv9 and beyond) will require new licensing agreements. New specific designs (Matterhorn/whatever comes after A78) will require new licensing agreements.

Future iterations of the Neoverse reference design and associated interconnects will require new licensing terms/agreements.

@SarahKerrigan

I don't think Apple will owe nVidia anything unless they sign new licensing agreements. The existing ARMv8 license should cover all their in-house CPU design needs, at least for awhile. And there's nothing stopping them from just staying on that ISA in perpetuity.

In Apple they must be really happy
Working with nvidia....
Feb2021 Apple presents the new altivec instruction set....

 

[Doug S]

My understanding is that the ARM licensing terms say they have to. Not to mention the licensing terms (and the rates!) are probally not perpetual. That's what I mean by an out, Apple has to have a plan, because they are not paying a dime to nVidia.

There's no way Apple would have agreed to a non-perpetual license, and accept the risk that ARM (or some future owner) could wait for them to become too dependent and then jack up the pricing 100x at renewal time. I'll bet it is either perpetual, or the term is so long that for all practical purposes it is. Even if ARM's other architectural licenses are shorter term Apple would have just said "how much of a lump sum payment do we need to pay to make that a perpetual contract?" and got it done. Only a fool would believe Apple would leave themselves exposed to a risk of this size. You don't get to be a $2T company by being stupid.

 

[Doug S]

Apple doesn't have a choice this time. It is just like Apple's relationship with Samsung. They will not attempt to avoid them to the point of hurting their own bottom line.

Exactly. Apple is certainly no friend of Samsung's, but they had no problem spending billions to buy OLED displays from them because at the time they were the only company who could deliver what Apple wants. People who want to place human emotions and motivations on a corporation just expose their complete and utter ignorance of the business world.

 

[Doug S]

I don't think Apple will owe nVidia anything unless they sign new licensing agreements. The existing ARMv8 license should cover all their in-house CPU design needs, at least for awhile. And there's nothing stopping them from just staying on that ISA in perpetuity.

Of course they will owe Nvidia something. Nvidia will be the owner of ARM, so Apple's existing contract will remain in force. They'll make yearly payments for the architectural license (somewhere it has been said that's $25M a year, but who knows if that's correct since it isn't like we can see the contracts) as well as per core or per chip fees for their designs. Obviously much lower than the fees for using ARM designed cores but they still pay something.

But who cares, they will owe Nvidia exactly as much as they would owe ARM if they continued to operate independently. Apple isn't going to make big changes to their business model to avoid paying Nvidia $25M plus a few pennies per chip or whatever it is. They spend a hell of a lot more than that per year with Samsung, and arguably have more reason to dislike writing them checks than Nvidia.

 

[arandomguy]

Exactly. Apple is certainly no friend of Samsung's, but they had no problem spending billions to buy OLED displays from them because at the time they were the only company who could deliver what Apple wants. People who want to place human emotions and motivations on a corporation just expose their complete and utter ignorance of the business world.

Apple's emotions are really just a projection of personal emotions.

The Apple commentary reminds me of years ago when the topic in vogue was whether or not Apple would choose Samsung or TSMC as it's foundry of choice. I remember someone (who don't think actively posts here anymore) that was in the industry wondered why from a consumer angle people seemed so emotionally vested on whom Apple picked, but essentially it served as a proxy battle for AMD vs. Nvidia (and also Intel at the time), which as we all know many of strong emotions on. It's also interesting to note the sentiment on Samsung and TSMC has since swapped.

For all the emotional hand wringing over this deal I wonder what the correlation rate is for those that are against it from a market competitiveness standpoint on the other hand are not only fine with AMD becoming the dominant x86-64 CPU leader and GPU leader but supportive of it?

There is no way outside of severe double standards involving mental gymnastics that one can be fine (and even supportive) of the x86-64 CPU and GPU situation with respect to AMD and Intel while taking the opposite stance here.

 

[Tup3x]

I wonder what NVIDIA will do when it comes to PC CPU market. I would imagine that they'd like to sell ARM CPUs for people in the future.

 

[itsmydamnation]

I wonder what NVIDIA will do when it comes to PC CPU market. I would imagine that they'd like to sell ARM CPUs for people in the future.

the quickest way NV can kill arm is to complete with its own customers.
if its customers are pushing into the PC space with ARM NV should do nothing.

 

[DJinPrime]

The biggest loser has to be Intel. AMD eating their lunch in the desktop space and now having a real serious threat from ARM in the server space. For NV to spend $40B, means they're all in. With NV's money and talent, you can bet that a high performance ARM CPU will be coming. It wasn't that long ago that all the servers used in big businesses were IBM and Sun boxes. Since all the popular programing languages support Linux, even Microsoft with .Net Core, there's nothing that really ties people to Intel servers like in the desktop world.

Should have merged with NV and let Jensen run the show back in the days.

 

[NTMBK]

I don't think NVidia is going to directly compete in the phone market. They never had a credible 4G modem, and they certainly don't have a 5G modem- and without a good modem they don't really have a product that can compete with Qualcomm. Better to keep selling licenses to that market, plus try to sell them on NVidia GPU IP- they might have more success than Mali did. Surely the big prize would be getting Qualcomm to wind down their internal GPU and AI efforts and switch over to NVidia IP.

 

[DrMrLordX]

But who cares, they will owe Nvidia exactly as much as they would owe ARM if they continued to operate independently.

Interesting, I didn't know the licensing deal required them to make repeated payments to ARM. I guess NV gets that money now.

 

[KompuKare]

For all the emotional hand wringing over this deal I wonder what the correlation rate is for those that are against it from a market competitiveness standpoint on the other hand are not only fine with AMD becoming the dominant x86-64 CPU leader and GPU leader but supportive of it?

There is no way outside of severe double standards involving mental gymnastics that one can be fine (and even supportive) of the x86-64 CPU and GPU situation with respect to AMD and Intel while taking the opposite stance here.

I for one have only ever said for a truly competitive market, AMD needs to be on far more than 20% of units (and 1% of profits).
Aside from AMD shareholders, I don't think anyone wants AMD to be at 80% marketshare with 99% of the profits.
However, for a truly competitive market AMD need to be on more than 1% of the profit to be able to fund future R&D. Remember how back in the P4 days, Intel's vast profits allowed them to have a parallel design (Pentium M) which came to save them. Too small a share of the profits means not being able to afford a plan B...
That doesn't mean the profit split needs to 50:50 as in Intel's case their vast profits has allowed them to make very poor choices (Itanium, Larabee, McAfee, Atom contra-revenue, 4G modems), wasting $billions and not be affected.

 

[NTMBK]

Fujitsu hasn't ditched SPARC in the areas where they deal with Oracle.

Ah, interesting. I'm only familiar with the HPC side of things- thanks! I was under the impression that they weren't making a successor to the SPARC64 XII, and that's a 20nm part from 2017- felt like the end of the line for SPARC, given the ARM transition in their HPC line.

 

[ThatBuzzkiller]

There's a real risk of Chinese authorities blocking this deal just as it happened between Qualcomm and NXP ...

 

[Tup3x]

I don't think NVidia is going to directly compete in the phone market. They never had a credible 4G modem, and they certainly don't have a 5G modem- and without a good modem they don't really have a product that can compete with Qualcomm. Better to keep selling licenses to that market, plus try to sell them on NVidia GPU IP- they might have more success than Mali did. Surely the big prize would be getting Qualcomm to wind down their internal GPU and AI efforts and switch over to NVidia IP.

They could do whatever they want in desktop and laptop space though. Powerful ARM chip with GeForce graphics would be quite nice compared to Qualcomm's beefed up phone SoCs.

 

[Tabalan]

There's no way Apple would have agreed to a non-perpetual license, and accept the risk that ARM (or some future owner) could wait for them to become too dependent and then jack up the pricing 100x at renewal time. I'll bet it is either perpetual, or the term is so long that for all practical purposes it is. Even if ARM's other architectural licenses are shorter term Apple would have just said "how much of a lump sum payment do we need to pay to make that a perpetual contract?" and got it done. Only a fool would believe Apple would leave themselves exposed to a risk of this size. You don't get to be a $2T company by being stupid.

Oh yes, I'm sure everyone would love to cooperate with company which increases license fee 100x
I'm sure they'd have tons of clients lining up to ask for possibility to cooperate.
/s

 

[Tabalan]

The buyout of the year. My guesses might be:
- ARM CPU development won't be affected by now. But with the current situation some licenses will be stuck in certain version (ARM v8 or v9 if they made a negotiation before)
- ARM GPU development on the other hand seems to meet it's end after this generation. nVIDIA would introduce their own approach after said generation in order to win more money

- The licenses might be perpetual, but they might pay if they want to meet the standards of ARM. I see Apple and Huawei starting their own way to be apart from that standard. But for different reasons each one.
- Qualcomm, Samsung and everyone who has CPU custom cores won't be that affected at all. Well.. not that much for Samsung, they need to make their little cores custom too then if they want to go with AMD GPU
- However the ones who might be totally affected would be Mediatek, UNISOC, Amlogic and the ones who goes to the standard licence in order to save cost. It might be even mean the end of the road for them.

Even worse if the politics enters with full force and bans the chinese companies to renew the licence. This won't look well, since only Huawei seems to have the full licence to make a fully custom core from CPU to GPU.

ARM CPU development might speed up as Nvidia might want to fight in server market.
ARM GPU (Mali) was not competitive for long time, so drastic changes are welcome.

Qualcomm and Samsung don't have custom cores. Qualcomm tweaks vanilla Cortex A, while Samsung ditched it's custom CPU team and goes back to Cortex (with possible mild tweaking, like Qualcomm).
Nvidia will not burn 40 bln $ by making it impossible to renew ARM license, let's be real. Plus, from what I've read ARM licenses were increased recently.

Chinese already have problems, can't use TSMC, Samsung. So ARM ban won't change much.

 

[Hitman928]

There's a real risk of Chinese authorities blocking this deal just as it happened between Qualcomm and NXP ...

I don't see how that could happen at this point. The reason China could stop that Qualcomm/NXP merger was because Qualcomm had most of its revenue coming from China. That's not true for Nvidia and there is a separate ARM China subsidiary (where ARM only owns 49%) which has pretty much already gone rogue at this point. That along with the U.S. bans for tech IP for China, Nvidia has to be prepared to just cut out China from their business and move on leaving China no power to stop the merger.

It is much more likely that U.K. regulators may step in but seeing as ARM was already owned by a Japanese company, I don't know if they actually will or not even though Nvidia could be seen as a competitive takeover compared to Softbank's acquisition just being a portfolio expansion. We'll find out soon enough.

 

[moinmoin]

Well...

NVIDIA to Acquire Arm for $40 Billion, Creating World’s Premier Computing Company for the Age of AI

NVIDIA and SoftBank Group Corp. today announced a definitive agreement under which NVIDIA will acquire Arm Limited from SBG and the SoftBank Vision Fund in a transaction valued at $40 billion.

nvidianews.nvidia.com

Relevant quote:

Additional Transaction Details
Under the terms of the transaction, which has been approved by the boards of directors of NVIDIA, SBG and Arm, NVIDIA will pay to SoftBank a total of $21.5 billion in NVIDIA common stock and $12 billion in cash, which includes $2 billion payable at signing. The number of NVIDIA shares to be issued at closing is 44.3 million, determined using the average closing price of NVIDIA common stock for the last 30 trading days. Additionally, SoftBank may receive up to $5 billion in cash or common stock under an earn-out construct, subject to satisfaction of specific financial performance targets by Arm.

NVIDIA will also issue $1.5 billion in equity to Arm employees.

NVIDIA intends to finance the cash portion of the transaction with balance sheet cash. The transaction does not include Arm’s IoT Services Group.

The proposed transaction is subject to customary closing conditions, including the receipt of regulatory approvals for the U.K., China, the European Union and the United States. Completion of the transaction is expected to take place in approximately 18 months.

• So transaction is expected to take 18 months, thanks to all the necessary regulatory approvals. This story will go on for that long until there's really a clear view on the impact of it.
• The transaction doesn't include Arm’s IoT Services Group, which was the part SoftBank wanted from Arm to begin with.
• With $21.5 billion in NVIDIA common stock the majority of the price is payable in stocks. The number of NVIDIA shares to be issued at closing is 44.3 million, that makes SoftBank the third biggest holder of Nvidia stocks, over 7% (see e.g. https://finance.yahoo.com/quote/NVDA/holders?p=NVDA)
• The cash payable is limited to $12 billion, starting with $2 billion for the first installment.
• NVIDIA will also issue $1.5 billion in equity to Arm employees.
• The remaining $5 billion are subject to satisfaction of specific financial performance targets by Arm and may be paid out in cash or common stock under an earn-out construct.

Edit: Here's Nvidia's presentation from this morning: https://s22.q4cdn.com/364334381/files/doc_presentations/2020/09/NVIDIA-Acquires-Arm-FINAL.pdf
Also a webcast link can be found on https://investor.nvidia.com/events-and-presentations/events-and-presentations/event-details/2020/NVIDIA-to-acquire-Arm/default.aspx

 

[moonbogg]

the quickest way NV can kill arm is to complete with its own customers.
if its customers are pushing into the PC space with ARM NV should do nothing.

I don't understand this. What do you mean by compete with its own customers? What customers? ARM buyers or GPU buyers? The idea is Nvidia would sell ARM CPUs for desktop use. How is that an issue?

 

[Doug S]

I don't understand this. What do you mean by compete with its own customers? What customers? ARM buyers or GPU buyers? The idea is Nvidia would sell ARM CPUs for desktop use. How is that an issue?

Qualcomm is already trying to create an 'ARM PC' market, so if Nvidia sold chips targeted for such a market they would be directly competing with one of their largest licensees who is hoping to expand into a new market.

If on the other hand Nvidia cares primarily about servers for the market they hope to enter, they wouldn't be stepping on the toes of any major licensees. There are a few tiny players with big hopes and dreams but not much scale to speak of, and Amazon who cares only about highly customizing chips for their own use so isn't a customer Nvidia could win anyway.

 

[Saylick]

So if the goal was to create ready-to-license ARM servers with Nvidia IP built-in, was $40B worth the price of admission? If I'm not mistaken, ARM doesn't make that much money off of licensing business (maybe even operating at a loss), so the addition of NV IP into license-able products needs to be super profitable in the long-term to make the entry fee worthwhile, no? If Nvidia jack up the price of its licenses for the IoT and mobile processors, that will disincentivize their own customers from staying with ARM, so I don't see that as likely. This leaves ARM+Nvidia developing an in-house, state-of-the-art ARM server processor with high licensing fees, which I thought could be developed already by Nvidia as they currently have an ARM license. In other words, is ARM IP and owning its in-house engineer team worth $40B? Nvidia must have thought so, and I guess the price is easier to bear since Nvidia stock is pretty much at an all-time high.

 

[coercitiv]

If Nvidia jack up the price of its licenses for the IoT and mobile processors, that will disincentivize their own customers from staying with ARM, so I don't see that as likely.

What exactly are current chip-makers going to do?! Spend years and billions to (separately) create new hardware and software ecosystems that may or may not succeed at competing with ARM? Let's build a brand new standard and compete from scratch with mature ecosystems such as x86 and ARM.

It will only take them a decade.

 

[Doug S]

So if the goal was to create ready-to-license ARM servers with Nvidia IP built-in, was $40B worth the price of admission? If I'm not mistaken, ARM doesn't make that much money off of licensing business (maybe even operating at a loss), so the addition of NV IP into license-able products needs to be super profitable in the long-term to make the entry fee worthwhile, no? If Nvidia jack up the price of its licenses for the IoT and mobile processors, that will disincentivize their own customers from staying with ARM, so I don't see that as likely. This leaves ARM+Nvidia developing an in-house, state-of-the-art ARM server processor with high licensing fees, which I thought could be developed already by Nvidia as they currently have an ARM license. In other words, is ARM IP and owning its in-house engineer team worth $40B? Nvidia must have thought so, and I guess the price is easier to bear since Nvidia stock is pretty much at an all-time high.

Reports are that ARM makes about a half billion dollars a year. In other words a P/E of around 80, which is not far from Nvidia's overall P/E. So I guess they at least will fit in with the rest of the business as far as being overvalued.

 

[Saylick]

What exactly are current chip-makers going to do?! Spend years and billions to (separately) create new hardware and software ecosystems that may or may not succeed at competing with ARM? Let's build a brand new standard and compete from scratch with mature ecosystems such as x86 and ARM.

It will only take them a decade.

Are you saying that you're in favor of opportunistic price gouging then? Why not have Nvidia jack up the licensing fee by 10x because it's all about making money and screwing the customers, right? Martin Shkreli would be proud.