Nvidia prepares to abandon $40B Arm bid
bloomberg.com
bloomberg.com
NVIDIA being a well known hostile company, the industry did not miss the news about its ARM purchase intent. And thus they looked for alternatives. That's RISC-V.
Today, pretty much every company designing microcontrollers or SoCs is involved with RISC-V. They won't cancel those efforts to embrace ARM again.
New players who are looking for a long-term strategy are looking at RISC-V (as you note) in part due to Nvidia's reputation. Despite the deal being quashed, they're probably going to continue along that route.
However, it seems like ARM licensees might also be making plans to become less dependent on ARM (the company). The Nvidia/ARM deal was announced September 2020. Qualcomm announced that it was going to buy Nuvia (founded by ex-Apple Silicon people) in January 2021 along with plans to launch their own internally-designed CPUs (rather than relying on ARM reference designs). I think this is a smart plan regardless of the Nvidia/ARM deal, but creating your own cores means that ARM has little-to-no leverage over you. You don't need their new CPUs. Sure, ARM still updates their architecture periodically, but you can happily continue along making your own CPUs under a perpetual license. Qualcomm is making the jump to its own cores so that it doesn't need ARM reference designs.
SoftBank is somewhat souring on ARM since it isn't getting the huge returns it was expecting by owning the CPU that everyone was going to use in the future. Will SoftBank end up short-changing R&D at ARM making their reference cores less competitive with Apple Silicon and future Qualcomm cores?
Even if SoftBank was able to sell ARM for $40B, they bought the company for $31-32B in 2016. Basically, the sale meant only 5% annual returns which is pretty bad compared to the huge gains an index fund would have made during that time. I think there's going to be pressure to find a way to squeeze money out of ARM.
Heck, with Qualcomm moving away from ARM reference cores, that's a lot less revenue for ARM. Even if Qualcomm cores aren't better, Qualcomm is still a huge portion of the mobile CPU market. Every ARM CPU maker that starts making their own custom cores means less revenue for the reference designs and less reason for ARM to invest in those reference designs. If ARM starts reducing investment in their reference designs a bit, that will put pressure on more companies to create their own custom cores.
The sky isn't falling and I don't want this comment to sound like that. I think ARM has a fine, sustainable business. At the same time, it seems like there will be pressure from SoftBank given the lackluster returns even with the Nvidia deal and Qualcomm buying Nuvia puts further pressure on ARM and it's hard to justify a major investment in something that is likely to continue having lackluster returns.
If the rumor Softbank wants to IPO ARM is correct, there will be pressure to get all the fundamental metrics of the business pointing up and to the right to maximize the valuation of the IPO. The increased proceeds from a good ARM IPO vs just an 'okay' IPO would likely swamp any potential cash that Softbank might be able to milk from ARM in the short-term.
It will be interesting to see how much of ARM Softbank intends to offer in the IPO vs how much they might continue to hold.
I think an IPO might be the only exit strategy Softbank has for ARM.
A European company with lots of money. China will make noise if a US company wants Arm.
Arm will be acquired by Volkswagen, you read here first !
Apple will probably mess with it for 10 years like they did PowerPC until industry comes up with something totally new and revolutionary and then we will see a convergence on consumer devices across the board to whatever that is.
You talk about the $30b case as if a retail is taking higher risk/return portfolio but putting the entire $30b is not the same as buying something that looks promising with your phone.
If you can't fill the $30b into an asset, it's at 0%.
Everyone that wants an arch license has one (Intel, AMD, Apple, QC, Mediatek, Broadcom, Marvel, Huawei, Samsung, etc) and doesn't have to put down cash, they can just poach ARMs best chip designers away from them. ARM gets to strangle the $10 landfill router market until RISC-V takes over.
Apple aside, laptop and desktop sales are still x86's to lose, so unless both Intel and AMD take major stumbles, I dont see ARM shipping in much more volume in PCs than it does now.
If Intel/AMD can achieve even half of Apple's power efficiency on laptop/desktop ARM isn't going anywhere on client anytime soon.
It's hard to survive a bodged takeover attempt, much less an actual takeover.
On the other hand, RISC-V seems to be used so far primarily as a stalking horse to keep ARM in line.
In terms of "if the deal fails", yes. If the deal succeeds, they made out like bandits.
Whilst this is a bit overdone, it does highlight the challenges that Arm faces. If the Nvidia deal is dead - which seems likely - then floating clearly seems unlikely to offer the prospect of the return that SoftBank was expecting when it paid a premium for Arm.
The key question from an Arm user / customer’s perspective seems to be ‘Can Arm as a stand-alone company finance the investment it meets to stay competitive?’
[1] https://assets.publishing.service.gov.uk/media/61d81a458fa8f...
> As Arm’s CEO, Simon Segars, explained: “We contemplated an IPO but determined
> that the pressure to deliver short-term revenue growth and profitability would
> suffocate our ability to invest, expand, move fast and innovate.”
It sounds like Arm needs a change in leadership to me. Find the capital for long-term investment somewhere and follow a path to improvement like AMD did. Yes, there is pressure to deliver short term profit as there always is today. However, that's not a strategy that works for a company like Arm. Everyone around the world can see the value in what Arm produces. Find someone to invest who isn't a hedge fund manager and still sees that value.
They did, it was called SoftBank and now SoftBank needs its money back. The comparison to AMD isn’t comparable; AMD already had a high margin business in a proven market; it’s unclear what ARM is going to be pressured to do; either they will have to put the screws on their licensors or start making their own chips (which would be no different from the concern under nvidia)
https://semianalysis.com/the-semiconductor-heist-of-the-cent...
Nonetheless Arm seemed to do OK as a public company for 25 years but Son thought he saw an opportunity and paid a big premium for it.
I do think there is an issue around Arm as a relatively small player competing for talent with Apple, Intel etc which feels suboptimal for the industry as a whole. Perhaps eg the cloud hyperscalers could jointly and directly fund development of server chips in the same way that firms prepay TSMC for capacity.
Softbank tends to have a habit of overpaying. Maybe they just have more access to capital than available opportunities they can intelligently spend money on.
You can lose money in a subsidiary so long as it is small by comparison to your overall P&L (Waymo losses of a few billion over several years vs Google annual revenue is over $160B and profit over $34B).
If on the other hand, Waymo were its own public company (and not a subsidiary), then it would need to show results on its own.
Don't get me wrong: being a subsidiary of a public company is still worse than say being a private company with a massive warchest or being the subsidiary of a private company with strong cashflow / cash reserves.
The markets more want you to be honest in your 10-k and S-1 than strictly require short term profits.
* Do they each benefit from the R&D done by ARM?
* Is that benefit enough to be worth the cost, even shared amongst the recipients?
* Can they "agree" on an arrangement, and can ARM function well enough to stay healthy?
They will turn around that statement on a dime and not a single investor will blink with "...but you said to the regulators...". No risk in making such a silly statement.
They probably will get a change of leadership though. Often they chose the best-man-for-the-merger, and when it doesn't work out, they then actually search for the best-person-for-the-job.
They apparently want to sell the company and benefit from it together ... so it this surprising?
https://www.barrons.com/articles/qualcomm-offers-to-rescue-a...
I wouldn't be surprised if Arm employees create more value for other companies than they do for Arm.
You expect an open consortium to continue to innovate in an industry leading way? An open consortium is how you kill ARM. ARM didn't get where it is now by being sub-par.
I appreciate your intent and it seems like a nice idea, but my experience with community development means ARM would turn into an absolute shit show as various stakeholders fight for control and ram through a mish-mash of their pet personal features.
There's also prior art in RISC-V
Then when they have the chance to almost completely start over… the make BLE with approximately 1/2 of what people want and spend years bloating the spec on that too.
I’ll take a talented lunatic with autonomy over 20 different voices in a bureaucratic community every time.
the historic alternative is a slightly lower cost (most of BT complexity is of the plumbing type, not the PhD type) and much more significant market risks (due to incompatibilities: smaller addressable market and less ability to pivot from one market to a different one).
from the engineer PoV, i agree with you: i’d love for these specifications to be simplified. from the business PoV, i don’t think it actually matters that much.
(I realize there's some ongoing development costs there - that $20b a year is conditional on you spending $500m a year to make next year's game - but there's also no reason to doubt that revenue stream will continue either. COD is shovelware (it's AAA but you get my meaning) and if you just don't mess it up (BF2042 says hi) then there's no reason to doubt the customers will keep coming back. And again, that's 2.5y revenue of just one asset of a giant company. Cheap cheap cheap.)
Who gets the money?
"Two weeks ago, at the Code Conference, Endeavor CEO Ari Emanuel claimed “the total addressable market of content is infinite.” Netflix is spending $17 billion a year to validate his thesis. So far, they’re both right."
EDIT: That's not to take away from what architects (ISA designers) do. It's extremely difficult to develop precise and consistent ISAs especially when things like memory ordering, interrupts, pipelined execution, TLBs, etc come in to play. And they are gradually gradually adding features and improving. But this is not value to the tune of over half a billion dollars per year just for ARM licensing (then imagine x86 equivalent numbers for that very locked in ecosystem).
ARM is not in a rosy place with their ISA. Their biggest customers like Apple can and will lift and shift their entire ecosystem if ARM do what it's told. Android could start supporting other ISAs too if someone was motivated enough. And many of ARM's new customers outside the traditional low end embedded world (e.g., in servers) who have been the first ones to move to ARM have done so because their software stacks are very portable. This means they can relatively easily move off ARM as well.
Their logic designs for high performance CPUs don't seem to be top tier either. Not to say that in a disparaging way but for example the PA Semi team Apple bought blew past ARM's high performance mobile designs with their very first ARM core. https://www.anandtech.com/show/6330/the-iphone-5-review/
ARM is pretty decent but behind the likes of Intel, AMD, Apple for high performance cores. So if the ARM market gets bigger and more lucrative especially in the server space, they are vulnerable to other companies selling their own chips and cutting ARM's designs out of the high margin sales. Amazon could be designing their own core right now to use in their ARM servers in a few generations for example. Just as Apple did with its iphone chips. AMD could start making ARM server CPUs if that's where the money is.
Would require an architectural license and lots of talent and time.
Getting stock Neoverse cores is an incredible deal for them, this is probably a big part of the reason they can offer cheaper services from their chips.
Arm cores aren't really all that behind, there are just tendencies for wildly incorrect comparisons (past Arm cores vs upcoming competition cores, scale-targeted configurations vs workstation-targeted ones, etc.)
Of course, the ARM tax. Is there any reason why that would be a problem for them to get?
> and lots of talent and time.
Well just more talent than ARM Ltd.
> Getting stock Neoverse cores is an incredible deal for them,
What's the incredible deal? IP licenses are the other part of ARM's business model.
> this is probably a big part of the reason they can offer cheaper services from their chips.
The problem is when other companies buy the exact same core or customers need single thread performance other designs offer, and then Amazon decides they can make a better design themselves. Like what Apple did.
> Arm cores aren't really all that behind, there are just tendencies for wildly incorrect comparisons (past Arm cores vs upcoming competition cores, scale-targeted configurations vs workstation-targeted ones, etc.)
Traditionally they have certainly been. It was a huge embarrassment that the small PA Semi start up could design their PA6T core in a couple of years from scratch and introduce it at the same time as ARM's A9 (which was a venerable and relatively efficient but pretty poorly performing core considering they had big mobile, PC, and server ambitions).
Things might finally be turning around very recently just with their latest core, but it's difficult to know yet.
So really I think the synergy between microsoft and activision is where a lot of the price comes from as opposed to the actual cash flow that activision brings in.
Is there a RISC-V analog to Blizzard/Activision?
1. ARM doesn't own any factories. Its entire workforce publishes blueprints for making chips (like a software company where the entire asset is intellectual)
2. Buying this type of intellectual asset, means owning and controlling a technology.
3. This also means, all the other customers who depend on this tech (Apple, Samsung, Amazon, pretty much all big tech companies) are now at a disadvantage with NVIDIA as a competitor.
4. China heavily depends on ARM (Huawei, the company's biggest tech manufacturer rely on ARM)
5. This means, the after ARM gets owned by a US company, the US can possibly just cut-off ARM supply to China
6. Since ARM is basically like a software company, it's better to not be owned by a hardware maker. That way, it can prioritize demand from several hardware makers instead of being directed to cater to one market)
So, all-in-all this is a good thing :)
(2) "controlling tech" All big players: Apple, Nvidia, Samsung, Amazon, Qualcomm, Intel, .. have so called Architectural license with heavily crafted clauses in them that make them free from Arm control except for some minor details. They use just the instruction set and make their own microarchitecture.
(3) Arm China was robbed from Arm. The CEO stopped taking orders and just kept IP and is running company like their own. Chinese courts did nothing. China does what it wants inside China.
(5) No difference. There are too many cross-atlantic IP and design tool connections. Arm must comply completely to US government sanctions.
(6) Just like Nvidia. Both fabless hw IP companies. Difference is that Nvidia sells chips, Arm sells IP. Nvidia wanted Arm because they want to sell Nvidia IP to others. Nvidia has Arm architecture license, they don't need Arm IP to use Arm.
Did ARM cut off any new IP access from ARM China? Seems like a logical next step. And how is ARM China going to continue innovating? On their own?
Why not?
When I was young, the common wisdom was that Japan couldn't innovate, and once I saw an old bit of early 20th century analysis that said Germany couldn't either.
Both of those were wrong, why is China so different that it would be otherwise?
Yes. Thery may do badly, but this isn't going to be much of a problem for China for ages. They are unlikely to do that badly though - among 2 billion people there will be plenty of good people.
Their much bigger problem though is their lack of access to cutting edge semi-conductor manufacturing technology. I imagine they are on this though, probably through industrial espionage (invading Taiwan would help, but they will face similar issues for acccess to tech long term unless they also get access to ASML work I think).
The MIPS processor was copied for production in China (illicitly, until fully licensed), as was the DEC Alpha. There is significant processor design knowledge, and ample ability to copy any new designs produced by ARM-UK, even if they have to be scaled up to 14nm for domestic production.
Oddly enough, I learned recently that Russia prefers SPARC (known as Elbrus).
Arm China is now AnMou Technology. They have "two wheels" strategy. ARM CPU architecture development for local customers and new Core Power architecture.
> There is an ongoing dispute between ARM and ARM China.
>
> But the accusations that ARM China had stolen ARM IP and was relaunching it under its own banner? Those don’t appear to be true.
[1] https://www.extremetech.com/computing/326617-arm-refutes-acc...
Technically Soft Bank still owns 49% of arm-china subsidiary, and they could theoretically take control back... but the belligerent CEO of arm-china has supposedly setup private security forces barring access to the facilities, and there is something about a legal seal. So far Soft Bank has opted not to escalate the situation any further. There is of course a lot more to the story.
for those not aware: seals are a big cultural thing in asian cultures (at least Japan and China for sure). I saw some surplus processors shipped from china with a stamp on them, I asked a Chinese friend about it thinking it was some kind of disposition mark (trying to make sure it didn't say "DEFECTIVE" or something ;) and she told me it was the seller's name, and that was his personal stamp, basically like his signature.
That got me looking into it and a signature is a pretty good analogy. In Japan at least it appears you need a seal to do any sort of serious transaction (buying a house, etc). The seals are officially registered and indeed basically like a signature, if you stamp a document that means it's "signed".
For a business, control of the seal is pretty much control of the business, I'm guessing. It's certainly going to be difficult to do any governing of the company without it, even if you otherwise have legal ownership of the company it's going to be difficult to exercise it without the seal.
Bit of an interesting cultural touchstone, seems minor to westerners but it's apparently a big deal to them.
Qualcomm used to make their own (back in 2014 or so) but hasn't since. Samsung tried after and quit in 2019. The cloud cores are all standard ARM designs (neoverse).
Fujitsu has the fastest-ranked supercomputer with their custom ARM, and Apple felt confident enough in their M1 ARM to evict Intel.
A few players doing otherwise doesn't change that, regardless of how well they execute.
And many of these "completely resigned" CPUs are nothing more than small (but important) adjustments of the pipeline and memory subsystem.
Is that really the case? My understanding is that while, yes, they make their own microarchitectures, they rely heavily on IP from ARM to make that happen
Do they write their own instruction decoders, FPUs, etc? I thought they started with the reference designs for a core and then tweaked them to their liking, some companies tweaking more than others
A peek inside Qualcomm’s upcoming chips, for example: https://www.anandtech.com/show/17091/qualcomm-announces-snap...
All of the main cores are ARM reference designs. Qualcomm does add proprietary IP, but it is more oriented around their strengths, like integrating their 5G modem into the die, which is something that none of the other big chip manufacturers can do at the moment (to my knowledge)
Qualcomm did buy Nuvia though, so they might yet come back with something new.
Apple designs everything by themselves.
Certainly, they may choose to derive their implementations from ARM's designs (or use them wholesale), but the license allows them to make their own.
That said, in the mobile space, AIUI only Apple and ARM are nowadays developing their own implementations. In the HPC space, there are others (Fujitsu, Marvell, etc).
(ARM also sells more limited licenses which only allow their cores to be used.)
Lots of people calling this out as 'a bad thing', but at the end of the day the Chinese government/courts handled in what it thought was in the best interest of Chinese citizens.
For other countries that might look like respecting IP clauses but for China it doesn't seem to be. I think it makes perfect sense and is perfectly moral for a country to do so.
and for 6) I think one of the fears there is that nVidia would use ARMs near de-facto monopoly to force their tech onto the market and push out competitors, like qualcom on the mobile GPU market.
Whether it be trough integrating nVidia tech more deeply into the architectural offerings essentialy forcing competitors to license both techs, or by using the ARM IP to in the future outcompete direct competitors by charging more for the IP that they can now use without any cost. Even if they're not planning any of that, I think the fear that they might in the future is what's giving many people (and regulators) pause.
ARM itself is never in direct competition with its customers _because_ it only sells IP, nVidia sells chips and is in direct competition with others who depend upon ARM for their chips.
In the short-term, you're absolutely right. In the long-term, no one will continue to invest in Chinese businesses if China gets a reputation for banditry like this. Is it in the best interests of Chinese citizens to ruin their reputation for the next generation?
Of course you can do that but that is not how you do trade. Trade requires trust and a move like this undermines trust. And I find it difficult to argue that it's in the best interest of citizen to undermine foreign investors' trust in the marketplace. In the end it means that less money will flow in.
If somebody came into my house, ate my food, set themselves up in a bedroom and enjoyed the comforts of my household, then when I told them to leave declared everywhere they had lodged and dined in my house an independent territory, that would be an immoral act. Blatant theft in the eyes of anyone.
Allen Wu was removed from his post. Not only did he decline, he took off a chunk of the company with him. That is a move of douchery in business I've never seen anywhere in the Western world in my time alive.
If the company I work for fires me, I will leave the premises. I may not like the decision but I respect it. And in Denmark we have courts of law that ensure one vacates the premises by the date of termination.
The CCP is playing fast and loose with whatever it likes. That's bad behavior, whether you're a business or a human being.
China has already hijacked the ARM branch in China[1] and taken over ARM's IP.
[1] https://semianalysis.com/the-semiconductor-heist-of-the-cent...
Yes, ARM China didn't steal any ARM UK's IP. But ARM China is also no longer under the control of ARM UK, practically speaking. And the New IP offered by ARM China are also independent of ARM UK. I am wondering if the deal with ARM China and ARM UK are the same as AMD's JV, where China currently has AMD Zen's IP. Given the people involved I would not be surprised.
ARM UK are also well aware of the RISC-V threat, which China is currently pouring all the resources into it. I would not be surprised if you see a free high performance RISC-V IP offered by China just to destroy the ARM market along with some other x86 market. The threat is real. But then again HN will rejoice because it is free and RISC-V.
https://riscv.org/news/2021/10/alibaba-announces-open-source...
New Zealand too .....
It's is not. Being a fabless designer of hardware IP doesn't make them "basically a SW company".
It's still very much a HW IP company any way you slice it.
* Their work is in both cases a mix of design and engineering.
* Both produce goods that are reproducible at zero cost, and are protected (only) by intellectual property laws.
So while ARM definitely does not produce software, their work is somewhat comparable.
> Both produce goods that are reproducible at zero cost, and are protected (only) by intellectual property laws.
You forgot fabrication. Unlike SW IP where as long as it compiles it's ready to ship but stays virtual, ARM's IP must be manufacturable into physical things you can touch by the major fabs therefore must be grounded in the processes and cell libraries that those fabs can manufacture.
Unlike a SW company, they can't just freely innovate whatever shiny new IP they want without concern for the silicon manufacturing processes, therefore it's closer to HW than SW, as all their IP is eventually manufactured into real things you can touch and therefore must follow the manufacturing constrains.
If you follow their news announcements, they constantly talk about their partnership with Samsung and TSMC to adapt their IP to each of their upcoming process nodes, so their their customer like Apple or Qualcomm can just buy the IP and plop it into their design knowing it's already been validated for Samsung/TSMC and can reliably be sent to the fab. So ARM is still very much a HW IP company.
I mean the blueprint is zero cost but building your own fab costs tens of billions of dollars these days.
I would guess there are a lot more ARM licensees than there are fabricators. Most licensees probably pay somebody else to manufacture their designs.
i thought it worth mentioning
I don’t see the distinction? NVIDIA is fabless too??
(Heck, so is AMD I think)
And yeah, I’m not sure this distinction the poster is making between arm as like a software company and NVIDIA like a hardware company makes much sense.
Arm makes IP (in the chip design sense). NVIDIA makes IP and combines it with IP from Arm and others, and produces some product designs based on the chip, which NVIDIA tests and writes software for. But ODMs make the products and fabs make the chips. I really don’t think calling some of this like software and some like hardware is very explanatory, at least partially because NVIDIA writes a lot of software, and because I think the person reading that description might come away with the impression that arm just produces the architecture, and not actual core designs (which NVIDIA, Denver aside, uses).
Global Foundries has sizable operations in Dresden, Germany. Interestingly, this was a major semiconductor supplier of the Eastern Block, prior the fall of the iron curtain. AMD placed their primary foundry in Dresden, likely for the infrastructure and technical knowledge.
Global Foundries has decided that more money can be made at 14nm and above, than what is required for smaller process nodes.
Good thing ARM China already decided to go its own way [0].
0: https://semianalysis.substack.com/p/the-semiconductor-heist-...
This doesn't differ too much from other nations except in degree and in their ability or willingness to hide that fact.
Of course, there’s no separation of powers in China, the courts are not independent, they are part of the Communist Party of China. Nor is there a constitution that all court decisions must uphold. The only guidestar of the Chinese court system is to keep the CPC in power, and uphold laws passed by the CPC, nothing else.
> This doesn't differ too much from other nations except in degree and in their ability or willingness to hide that fact.
It differs from Constitutional republics where the court system is independent from political parties and mandated to ensure all laws passed by the legislative branch and actions taken by the executive branch do not violate the constitution, regardless which party is in power at the time.
Yes different political parties will try to pack the court when they can, but that pendulum swings back and forth over time. There’s a social contract with the citizenry that doesn’t exist in Communist countries - adhere to the constitution or be kicked out of power in future elections. You can see the evidence in how often power changes hands between parties.
It’s a problem when those executives go rogue and there are many cases of it causing problems for companies in China, both for foreign and domestic companies.
This means that in practice the US can already cut off supply to China.
See for example ASML: they are a Dutch company. So the US government only needed a friendly word with the Dutch government for the Dutch government to ban ASML from exporting certain advanced processes to China.
For the British government I'm sure there would be no need to call... A text would suffice ;)
(leaving aside all the drama with ARM China already because of those issues...)
If I recall correctly the parents used by ASML are owned by the US military so the US is part owner of the company -
But IP is very hard to control the supply of, especially one like this where hundreds of companies have a copy of the IP. If the US won't license it on fair terms, China will just stop enforcing IP laws and allow anyone to copy it for free.
What would happen is another mutually assured destruction stalemate.
NVIDIA isn’t a hardware manufacturer.
ARM's is to create a fair, competitive playing field between CPU producing companies and reap licensing fees for the pleasure, Nvidia's is to nakedly leverage all of its technologies for Nvidia's products' benefit and Nvidia's products' benefit only.
I don't think NVidia would be quite that stupid; I think if they were smart they would keep ARM business as usual just to avoid that, and just reap the licensing fees from the above companies.
Nvidia licensing their GPU to ARMs customers will itself be a big start.
https://www.ftc.gov/news-events/press-releases/2021/12/ftc-s...
Uh, exactly what have Google and Microsoft been up to? I'd say they're "too powerful" at this point...
Buying chips from Nvidia to power their data centers, for one.
(and if you can't sell assets, that's the only valid metric for valuation - "I need that thing and would have built it anyway". Even if it's complimentary to your existing business - the BATNA is "build it myself".)
extremely controversial (if not outright incorrect) statement to be labeling as "traditional wisdom". Says who?Probably not any economics textbook written in the last century.
I didn't write anything controversial and you don't need to consult an economics textbook, a basic Investopedia article will do:
From Investopedia:
https://www.investopedia.com/articles/fundamental-analysis/1...
"The dividend discount model (DDM) is one of the most basic of the absolute valuation models. The dividend discount model calculates the "true" value of a firm based on the dividends the company pays its shareholders."
An alternative model if there is no dividends:
"the Discounted Cash Flow model uses a firm's discounted future cash flows to value the business. "
The ability of a firm to sell itself to a foreign firm is not a factor is valuing the firm.
I wouldn't use these models to value a Pikachu trading card, but this is stock market 101 type stuff.
https://www.bloomberg.com/news/articles/2022-01-25/nvidia-is...
I wouldn't say that things are so bad for Nvidia. They are selling more cards than ever, their "Ultimate Play" to rise the prices across the board will, most likely, be successful in the long run. Sub 200 dollars segment is dead, like sub 100 before it. Since they are going to make monstrous MCM GPUs, the price for an absolute high end is going to rise to unseen heights(performance will rise with them). I hope Intel will stop this madness eventually, but they won't be able to do it fast, even if they try as hard as they can.
The salad years of buying almost anything at Walmart, Tesco, Amazon for 15% less than what it'd cost from the west and taking that to the hoop as success are long gone.
The low fruit is increasingly picked for them too. They can't do cheap labor 24/7/365 forever just like we can't do cheap prices for ever. Eventually we'll both need to rebalance price, jobs, security, and so on.
Not according to the vast majority of consumers. Look what trouble far less than a 15% rise in costs due to inflation is doing to consumers.
>They can't do cheap labor 24/7/365 forever
The west cannot do expensive labor forever. It's dying faster than cheap labor.
Picking a current tangential issue and globbing it on to a needed global trade correction which has been in the making for the last 20 years is opportunistic. Inflation like this hasn't been seen in the US since the late 70s, for which there are many factors including Covid.
>The west cannot do expensive labor forever. It's dying faster than cheap labor.
US labor & regulation (where I am) is medium. Not stupidly high certainly not lowest. I think other places in the west find it harder here than us. Still I am not particularly pleased with US governmental institutional competence in the last 15 years. We are nowhere near our best, and sadly sucking in a few important areas.
I would remind that outsourcing to China did not start the behest of a US politician taking a call from Joe Smoe manager who asked if he should outsource. The move east with consequent impact on labor and larger issues was done by US business people. Further examples are obvious, but let's not dismiss self interest either.
Going back to my original point: I absolutely could not care less what the Chinese think on this particular issue. ARM isn't theirs anyway. China, whenever strategic dependency arises, wants their own lock-stock-and-barrel-stuff. MC/VISA will never be top there. And Intel/ARM/AMD/TSMC will never rule the roost there either. They will not be dependent on the west. So my question for the west: how much and how long are you gonna play a part in their goal? What are the limits? How do we know when enough is enough?
Yes and no. China can't, no, but history tells us exactly what will happen. Japan used to be the "cheap and crappy electronics maker". At some point the population has been lifted enough that some other poor country will take over the bottom jobs. Likely somewhere in Africa. It is not as if China at some point will start to fall back to where it once were and even if it did all it would do would be to hold the production in China instead of the next country.
>Eventually we'll both need to rebalance price, jobs, security, and so on.
That will never happen.
I never said they were. Also babysitting Chinese outcomes is another not-my-problem. There's a whole invasive inflexible CCP state control thing going on there who, by the way, claim it as their problem.
It's also not a feature for China to fall back to some pre-Deng era. Knowing what I care and don't care about it is not equivalent to hoping China fails. Certainly not. Maybe China will rock and roll. Maybe they fall in disinflation like Japan did starting somewhere around the 1990s, from which I'm not sure they recovered.
I am saying the west has got to do a better job at realizing what the end game is for China in some important areas. And know what our concerns are, and do a better job at holding the line. Here, in the US, "show me the discount! show me the cheap!" cannot be the last thing heard.
>Eventually we'll both need to rebalance price, jobs, security, and so on.
Good! I love it when people lay down a hard line. Hard lines are easy to bust.
They're already adjusting their economy. The cheap labor is now found in the rest of SE Asia with east Africa viewed as growth potential for cheap labor by the Chinese.
https://www.hpcwire.com/2021/11/24/three-chinese-exascale-sy...
or
actually state protecting it?
When ARM was bought by SoftBank the main argument for it being acceptable was that SoftBank were not chip manufacturers in their own right. And that because of this, ARM would have every incentive to carry on treating all of ARM's existing customers reasonably fairly under SoftBank's ownership.
That's obviously not the case with Nvidia.
So while, it's possible to imagine that Nvidia wouldn't (at least initially) abuse any ownership of ARM. From a societal pov, why risk allowing a situation that relies on humans behaving well on an ongoing basis not to fail? And two, given the way ARM operates i.e. all customer $$$'s are equal, why would Nvidia bother if they didn't think there was an advantage?
Longer term the market might be in a better position with a public ARM.
I think a lot of the concerns around NVIDIA were driven by hyperbolic statements and memes.
ARM is a good example of a monumental success and relative failure as in their cores and their architecture is in everything but they never have been able to capitalize on their success at the same scale at least as far as it goes for converting it into revenue which sits only at around $2B per year.
NVIDIA on the other hand has mastered capitalizing on every success no matter how small it is.
Overall as acquisitions go on paper it was a good match. It would’ve given ARM access to a lot of capital, engineering resources and a driven management with a strategic vision.
NVIDIA would’ve gotten the ability to set the path for one of the most commonly used CPU architecture and being able to offer a completely vertically integrated solution to both enterprise and end consumers.
Arguably NVIDIA could achieve the latter on its own too at least by licensing ARM I do think it’s a shame that they don’t have an x86 license too. ARM on the other hand would need to undergo a massive change to get into the position they arguably deserve to be in.
ARM got to where it is by not being too greedy. If others perceived them to be making power grab then a lot fewer people would have been willing to stake their own futures on the architecture.
To mix metaphors, ARM went with a 'rising tides floats all boats' approach to grow the pie in general instead of just their own slice of it.
And the market now is more dependent on their reference designs and core IP than ever.
ARM as a company isn’t in a good position, they need a parent that can actual drive them forward or go public and get the investment they need and honestly deserve.
As in, they failed to destroy the company for short term gains, like companies usually do.
Considering where ARM designs and architecture are today their revenue and profits are pitiful and it does hold them back considerably.
ARM justifiably so should’ve been one of the largest companies in the world by now.
No way. Even losing the most popular electronics brand (Apple) didn't sway NVIDIA from its course.
> ARM on the other hand would need to undergo a massive change to get into the position they arguably deserve to be in.
Intel has completely botched the last six-ish years, and Apple proved that ARM-based processors cannot just compete with Intel's offerings but outright destroy them, while still keeping backwards compatibility. All that ARM (as an architecture) needs now is Microsoft also offering ARM support and runtime emulation in Windows and a CPU vendor willing to sell decently powerful chips to vendors... and then, snap, Intel is gone.
Err... plenty of places would be happy to have $2B revenue per year. ;)
Because it's only part of the problem. If Arm is so successful, it's (IMHO) because of software. Sure, they have world-class CPU designers. But in order to launch a new CPU, you need a full software ecosystem, as RISC V startups are discovering.
One other advantage of Arm is that it has strong anti-fragmentation measures in place. With enough money you can design your own cores, but in order to deviate from standard Arm architecture, you need Arm's signoff. This serves the first advantage: it keeps the software ecosystem value intact.
You should have seen the state Raspberry Pis were in circa-2011. Everyone online was treating it like the RISC-V of today, criticizing it for a complete lack of software and calling it a novelty board. Lo and behold, come 2018 everyone and their mother wanted a Raspberry Pi for some purpose. Sure, 70% of the software people wanted to use wasn't available, but the things it had were power efficient and performed just about on-par with it's x86 counterparts. RISC-V is between both of those stages right now, the biggest limiting factor is getting hardware into the hands of developers, which is starting to dissolve as manufacturers are catching on.
> This serves the first advantage: it keeps the software ecosystem value intact.
Why do people assume that adding an extension to your RISC-V processor throws the software ecosystem out the window? It's the exact same scenario as ARM, except you're not beholden to arbitrary version updates (eg. v6, v7, v8) that break compatibility. If you want to upgrade your ISA, you just... do. Your base instructions will still run fine, and software compiled for RISC-V will just run. The only way you could fragment like that is if your chip failed compliance tests, which... why would you even ship it then?
> You should have seen the state Raspberry Pis were in circa-2011.
Yeah, I was one of the naysayers initially. And in retrospect the biggest advantage of Raspberry was its price. It sold at a price-point where no one could compete, and that helped overcome most other disadvantages, in a self-sustaining snowball.
And that might very well be the case for RISC-V as well.
Except it’s not. A large RISC-V user could add their own proprietary extension that isn’t available to anyone else.
A large company could do it but do you really want to build your own LLVM backend? And the largest Arm customers do design their own CPUs.
Of course RISC-V potentially challenges this model.
How would smoothing out the current Intel/AMD machine not be healthy overall?
Nvidia should play ball (or somehow the regulators should nudge them to play ball ) and invite all big ARM customers to form a co-op.
Nvidia is a big contributor to the RISC-V standard, all Nvidia GPUs have had RISC-V chips in them for many years, etc.
So no need to wonder, this already happened, many years ago.
For example, Google's Tensor processor in the Pixel 6 uses ARM Cortex-X1, Cortex-A76, and Cortex-A55 cores and ARM's Mali GPU. They add some ML cores to the design, but the CPU/GPU is really designed by ARM. Apple, on the other hand, makes its own internally designed cores for its processors rather than using ARM's designs.
While you're right that the general public doesn't know ARM/Apple's deal, we do know that ARM offers a perpetual ISA license. Even if Nvidia bought ARM, Apple could still make current ISA processors forever (it seems unrealistic to think they don't have a perpetual license). While Nvidia might not want to help Apple, it would be in Nvidia's interest to offer new ISAs to Apple at reasonable rates because a) one probably doesn't need updates to the ARM ISA at this point b) Apple not getting on board with a new ISA could impact the rest of the industry getting on board with it given that Apple is so large (and respected) c) Apple puts a lot of time and money into LLVM and having them against an update to the ARM ISA would (at the very least) mean that there wasn't free labor (from Nvidia's perspective) adding compiler support for the ISA update.
It's true that we don't know all the details about deals between Apple and ARM, but at this point it seems like Apple doesn't really need ARM. Samsung, Google, Amazon, and others use ARM's reference designs. If ARM disappeared, they wouldn't get updated cores and would have to build up in-house design teams. If ARM disappeared, Apple would just keep on making new designs. I think Qualcomm is looking to go more custom in the future as they bought Nuvia and are looking to make inroads into things like laptops over the next few years.
In some ways, it seems like ARM getting bought by Nvidia would be good for Apple. If Nvidia becomes really harsh for third-party licensees, it could mean a few years where the costs skyrocket in the Android ecosystem while their costs remain the same. Even after that, it might lower the number of manufacturers for ARM processors. Would MediaTek spin up a custom-design shop? Would Samsung? Maybe, but it would add a lot of cost over re-using reference designs.
Apple isn't really reliant on ARM for anything at this point. The rest of the industry is pretty reliant on ARM's reference designs. If Nvidia ownership meant that those reference designs went up in cost or if Nvidia wanted its best work to go into Nvidia processors and put out weak updates, that would benefit Apple with Android manufacturers scrambling to figure out what to do: buy expensive Nvidia processors, ship weaker updates, invest in the custom-core route?
Arm and Softbank said so much last week.
Arm does not take profits from Apple CPUs.
https://webreader.app/?url=https://www.bloomberg.com/news/ar...
Now the question is, who will that "someone" be next? What kind of suitor is fit for ARM?
The curse of ARM being so successful and incredibly crucial to many pervasive industries, yet seemingly unable to go it alone.
They have GPUs. Got network capabilities with Mellanox. Add ARM knowledge on top and you have a complete platform building capability.
TL;DR: nVidia just wants to dominate the whole stack. Like Apple, but for data center / scientific / AI / HPC, etc.
HPC is nice, but when you hear Jensen getting really excited, it’s not about dominating some niche like that, it’s about a vision of the shiny sci-fi high tech future, and actually delivering the tech to make it real.
So don’t just look at HPC to understand the NVIDIA ambition. Start at edge computing; imagine a world with ubiquitous autonomous robots (cars and drones and otherwise). Think of the onboard chips driving their vision and speech recognition models: That’s a great place for ARM and NVIDIA chips together, whether as one company or two. Watch a recent keynote and see how all the rest of the tech fits into place as part of that: 5G signal processing chips, for instance, something you might gloss over if you’re not in telecom. You don’t need a roadmap to see how it is all connected in support of this world of the future.
(I certainly don’t have the roadmap, either, I just watch the keynotes and help shuffle bits.)
They seemingly don’t see any value in SW. architectures like end-to-end designs (DPU->Network->DPU->pcie) can be great but without SW to make them consumable it’s doa outside of dedicated clusters.
> Nvidia Corp. is quietly preparing to abandon its purchase of Arm Ltd. from SoftBank Group Corp. after making little to no progress in winning approval for the $40 billion chip deal, according to people familiar with the matter.
> Nvidia has told partners that it doesn’t expect the transaction to close, according to one person, who asked not to be identified because the discussions are private. SoftBank, meanwhile, is stepping up preparations for an Arm initial public offering as an alternative to the Nvidia takeover, another person said.
> The U.S. Federal Trade Commission sued to stop the transaction in December, arguing that Nvidia would become too powerful if it gained control over Arm’s chip designs.
> The acquisition also faces resistance in China, where authorities are inclined to block the takeover if it wins approvals elsewhere, according to one person. But they don’t expect it to get that far.
> Both Nvidia and Arm’s leadership are still pleading their case to regulators, according to the people, and no final decisions have been made.
Nationalistic flamewars are particularly nasty and particularly avoidable. No more of this, please.
If you'd please review https://news.ycombinator.com/newsguidelines.html and stick to the rules when posting here, we'd appreciate it.
We detached this subthread from https://news.ycombinator.com/item?id=30071505.
* When Chinese were ahead, European stole and smuggled manufacturing technologies from China.
* Industrial revolution in the United States was started by ruthlessly stealing the British innovations. Samuel Slater -- the "Father of the American Industrial Revolution" -- is known as "Slater the Traitor" in Britain.
* Japanese stole and copied every design and technology from the US and Europe they could from the start of Meiji era 1868- to 1960s.
Piracy is not a theft in a sense that it's just illegal, not fundamentally immoral. I think it's actually cool historical constant that moves the world ahead.
In as much as the Chinese firms (which in many cases are quasi-state-owned) that do the theft are breaking treaty and contractual commitments, then the piracy is both illegal, and immoral. Not honoring binding commitments is wrong.
I have personal experience with this, with, as it happens, Huawei, who licensed code from the European company I worked for, used it far more widely than the license permitted, and then, when we attempted to negotiate a broader license, simply dropped their license entirely, continuing to use the code until they had reverse engineered it and could generate new instantiations on independently of us. (That much I know to be true; I actually suspect, but don't have conclusive evidence, that employees of our Chinese subsidiary who went to work for Huawei stole source code on their way out the door, making most of the reverse engineering a simple hiring decision for Huawei. I also have reasonable evidence that another Chinese firm did the same with a major American technology company I later worked for, although again, it's difficult to prove).
[1] https://www.dw.com/en/chinese-exports-crushing-german-solar-...
[2] https://www.cleanenergywire.org/news/last-major-german-solar...
[3] https://www.npr.org/2020/01/17/797322305/episode-965-das-gre...
Many, many things were unregulated in the past but ARE regulated today.
If they wanted to maintain a policy of ignoring IP, etc., they they should not have joined the WTO and signed conventions that hold them to obligations.
These organization and conventions set the stage or provide the framework and law by which signatories are bound.
Your argument amounts to: hey, the US and Brazil and the Middle East and many other countries used slave workforces in the past therefore it’s okay for China to do the same today, else they are at a developmental disadvantage.
In any event, I'm quite sure any patent [had the concept existed] would have run out by the time they were adopted elsewhere.
Chinese companies are also aggressive in marketing stolen technologies but because of incomplete knowledge and expertise, end up in sub-standard/broken products, such as the capacitor electrolyte issues around 10-15 years ago.
Not disputing this but what technologies did Europeans ever steal from China? I’m only aware of tea and silkworms..
Piracy is definitely a theft. It’s ludicrous to suggest that IP doesn’t have any value like physical property.
Two wrongs make a right..
The point is it's typical of China to steal IP. It's a problem because it gives Chinese companies an unfair advantage as they grow big by stealing all they want from others, but have their own IP protected in other countries.
It would be nice to live in a world without patents, but that's not the world we live in. So if you're ok ignoring reality, might as well go on a tirade how stealing cash is "not immoral" because you don't believe in government issued tender or something.
China has done this to perfection in the real world.
Stealing a tech was a massive diplomatic blow in Civilization. The US just shrugged (because the rich made a huge amount of money and increase of relative wealth status by acquiring a massive slave labor force in China rather than deal with the uppity American middle class).
I just did a light google pass to see if any international relations academics have done anything with Sid Meier's Civ and various other types of games. I expected them to not, because of course academics are STILL "ew, computers" and even worse, it's gauche mass market entertainment.
But the abstraction is, I would argue, more detailed than a lot of academic analyses which are largely bloviation, the game theory quantifiable and measurable, and reduces a lot of complexity that normally would be hard for a garden variety person (aka a gamer) to wrap their head around.
Civ always tried dropping historical tidbits and education into it, but arguably its most potent contribution is simply the more honest treatment of history: civilizations rising and falling, fighting over resources, getting conquered, and getting destroyed, and the roles of economic strength, military strength, and tradeoffs.
It's not like anyone is denied property here. The conflict boils down to people doing things other people don't like / want them to do.
Once humans know how to do something, they know. It will spread, eventually becoming common knowledge.
Infringement is the right term here, and it's all about that spread, the timing, etc...
The US invented this practice, we didn't respect the IP of other countries until we started generating significant amounts of our own. China even stole this idea!
Comments both there and here.
Back in the day a Japanese businessman who screwed up the way Son did would find a knife in their room and know what they were supposed to do with it. The ARM deal took the pressure off but it is back on again.