SoftBank is buying ARM Holdings for $32 billion. Why would a company not presently in the semiconductor business spend 32 times sales to enter a new industry? By traditional measures it makes little sense. But for SoftBank it makes perfect sense, because here’s a company that has spent more than 30 years making high-risk bets on entering new businesses by apparently over-paying for assets. It’s the way they’ve always done it and it has nearly always worked. In this case SoftBank is paying a 43 percent premium over the recent ARM share price because that’s how much money it took to overcome the resistance of ARM management. And it’s just a guess on my part, but I’d say those very ARM managers are, for the most part, doomed. SoftBank isn’t buying ARM because Masayoshi Son is so impressed with ARM’s success: SoftBank is buying ARM because of what it can become.
The only way SoftBank can justify paying such a high price for ARM is if Mr. Son is about to change ARM’s business model.
The world is filled with fabless semiconductor companies — companies that design chips that are made in the factories of others. That such fabless companies exist is because it costs so darned much to build a fab — a semiconductor factory. So Intel has a fab and AMD no longer does. TMC and TSMC in Taiwan have fabs but no other companies in that country do. Yet the very expense of starting a fab makes it important to keep the factory running at 100 percent capacity 24/7 which explains why fabless chip companies can nearly always find someone to build their products.
But ARM Holdings doesn’t sell chips, they sell chip designs used by others to build ARM-based mobile processors. ARM customers include Apple, Qualcomm and Samsung — between them manufacturers of a majority of the processors used today in handheld devices of all types. These three companies use ARM technology to build the processors for hundreds of millions of devices each year worth hundreds of billions of dollars. And of the three, only Samsung owns a fab.
SoftBank covets the success of Apple, Qualcomm and Samsung. SoftBank covets even more the value of those companies — value SoftBank sees as coming from ARM intellectual property. “Why should they get all the money?” thinks Mr. Son.
What if ARM Holdings turned itself into Qualcomm? It wouldn’t take that much to do so. Both companies design mobile chips. Neither company owns a fab. It can be argued that most of the DNA in Qualcomm chips comes from ARM, anyway. Yet Qualcomm’s market cap today is $81 billion and SoftBank just got ARM for $32 billion.
What if ARM Holdings turned itself into Intel? It wouldn’t take that much to do so. True, Intel has a fab but there appears to be more profit these days in going fabless. Intel’s most profitable products today are server processors and ARM has just finished its first designs for those. In addition to dominating mobile processors, ARM could become a significant maker of server processors, too. Yet Intel’s market cap today is $164 billion and SoftBank just got ARM for $32 billion.
If we want to put a measure on Mr. Son’s aspiration in buying ARM, take the average of the Qualcomm and Intel market caps, which is $122.5 billion. By abandoning ARM’s traditional model of licensing processor designs to other companies and becoming a fabless semiconductor company, itself, Mr. Son believes he can turn his $32 billion investment into $122.5 billion.
And he can do it. ARM-based chips are dominant in mobile processors and entering new markets all the time. All ARM has to do is make and sell its own chips, which requires only hiring a new sales force. The company can continue in its existing business while starting to compete with its old customers because those customers are trapped for the next 2-3 product cycles at least — just long enough for ARM to steal their business.
Five years from now buying ARM Holdings for $32 billion will look like it was a steal. Apple should have done it. But then Apple doesn’t have Mr. Son’s killer instinct.
Interesting that Softbank has already been sponsoring Niantic’s Ingress virtual reality mobile game. After the rush to Pokemon Go, I think Softbank sees the future in higher end mobile processors to fully flesh out the augmented reality gaming model that will no doubt be the next genre that will take over the global gaming market.
Wow, Intel is dead. Avram Miller was right.
https://twothirdsdone.com/2016/04/24/intel-how-a-vain-of-gold-turns-into-a-big-hole/
This does not bode well for Windows Desktop on Intel.
If it means Microsoft dies, then I sure hope Intel dies… the sooner the better. There were so many potentially great microprocessor designs that died because Microsoft can’t or won’t port windows. Leaving us with X86 for so long even the rotting corpse smell is gone.
–
Linux really needs app support so the world can move on.
>It’s the way they’ve always done it and it has nearly always worked.
Didn’t Son and Softbank buy the old Seybold business and ultimately shut it down two or three years later? I guess that would be one of those times it didn’t work.
Also, didn’t Son and Bill Gates design the Radio Shack Model 100, the laptop computer with a one-line LCD screen? I loved that thing (though I loved the Model 200 more).
Kyocera designed/built and also licensed the Model 100 to Radio Shack.
https://en.wikipedia.org/wiki/Kyocera
The Model 100 had an 8-line LCD display, not 1-line. The firmware for it was the last, and IMHO best code Bill Gates ever worked on personally. 😉
.
It was also probably the first and only computer from Tandy/Radio Shack I’d have actually considered buying. 🙂 (The price premium for the bigger screen on the Model 200 was way too high.)
.
https://en.wikipedia.org/wiki/TRS-80_Model_100
And at the time, that little RadioShack computer gave corporations a significant boost in productivity by giving employees the ability to access email and product data in a portable fashion anywhere there was a phone line.
Interesting perspective, Bob, but remember that an awful lot of “best laid plans” for tech dominance 5 years out get pummeled. (Anyone remember WiMax? AOL’s dominance of the web?) More interesting to me is what this does to Sprint. The company stock was down about 6% today and Masa Son has already taken a $10 billion bath on his investment.
5 years ago Bob Cringely was saying that Blockbuster would be distributing movies on IPods.
Looks like that was your own prediction: https://www.cringely.com/2011/06/07/iclouds-real-purpose-is-to-kill-windows/ “This is Cringely’s new secret way to generate hit counts: Predict Apple domination. I’m still waiting for IPod to become the default video format for Blockbuster rentals.”
There’s a lot of resistance to buying desktop computers these days. I don’t trust Intel or AMD due to https://en.wikipedia.org/wiki/Trusted_Computing which doesn’t mention ARM. I’d buy a new desktop today with ARM and let the Chinese spy on me all they want, if it would keep the NSA et al out. And I don’t think I’m the only one to realize this. We’ll never know if Intel or AMD had their bizmodel torpedoed by National Security Letters.
> I’d buy a new desktop today with ARM and let the Chinese spy on me all they want, if it would keep the NSA et al out.
I wouldn’t make that trade. We know, or at least think we know, what the NSA is and what it’s doing. I won’t dispute that it’s pretty bad. However we don’t know what the Chinese are doing, where they’re going, or what their desired end-game looks like. I strongly suspect that if you’re not a Chinese national, or at least of Chinese heritage, in the long term playing “enemy of my enemy” them will not go well for you.
risc-v is the future.
Well, maybe…
ARM is easy to integrate and fab tested. If their technology gets too expensive, companies can hop to MIPs, risc-v or even some of the open cores processors. Becoming a fabless semiconductor company would make ARM compete with itself for end design wins. That is an ugly and difficult business.
ARM processors come from Samsung, TI, Freescale, Xilinx, Altera (Intel), and a dozen small players. There is no way they could restrict the licenses and serve the market. There are literally thousands of part numbers based on the ARM architecture.
The ARM sales team are whale hunters. They land a semi maker as a licensee, then add on training, consulting etc. The semi builder then finds the end users.
Winning end user designs is elephant hunting (Apple) down to deer and rabbit hunting. Their sales process will have to have a lot more people in the field.
SoftBank has to believe ARM has a good chance at the server market to make it worth the premium. If they believe the IoT hype, well they deserve to fail.
ARM has a level at which it can’t charge customers, because they will seek out an alternative of which many exist. Also charging existing customers more takes money out of those customer pockets, who aren’t going to be too keen on having less.
ARM also has the same problem Intel (and Nvidia) had and why Intel failed so miserably in mobile. The CPU is important, but the designs also need cellular connectivity. And cell tech is hard plus a patent minefield. The reason why Qualcomm is successful is because of the cellular side, which if you buy it also gets you a competent ARM CPU. The more functionality ARM adds, the more they compete with their customers and the less successful they will be!
Many years ago “multi-sourcing” was common in the semiconductor industry. Firms would agree to make each others products. This gave their customers more than supplier for the chips they needed in their products. Multi-sourcing encouraged cooperation, competition, and innovation. The early Intel CPU’s were multisourced. The Pentium was single sourced.
.
When you single source a vital chip, you can control the price and supply. With the rapidly growing PC industry, Intel could print money. When firms started designing new types of consumer appliances, they needed processors that were inexpensive and would stay that way. As they looked at chips for their game systems, cable and satellite boxes, etc Intel lost the bids. When smart phones and tablets came along the makers needed processors and once again Intel and their single source model lost out. ARM had a great processor and a much better business model. Firms could get the chips they needed at good and predictable prices, and from multiple vendors.
.
ARM’s current top line chips are very powerful. As new chips come to market, the old chips become much cheaper. The old chips are perfectly good and suitable for many applications. The Raspberry Pi 3 costs only $35.
.
The good folks at Softbank realized there will be more and more applications needing chips and their designs will be based on ARM licenses. The future is with ARM it was worth the premium for Softbank to be part of that business.
“Apple should have done it.” Don’t they still own a chunk of ARM?
Even if ARM remain fab-less, surely Apple will have to worry about increased licensing fees?
Apple invested in ARM in the late 80s/early 90s, and then sold their stake during the “Time of Troubles” before Jobs came back and made several billion dollars, basically saving the company. (at least, that’s how I remember the story)
Why would ARM want to compete with it’s own customers? I am sure ARM knows that most of it’s large customers already have a plan B in place in the way of their own home grown alternative to the ARM based chip, to de-risk exactly this sort of a misadventure from ARM.
Softbank can’t turn ARM into a Qualcomm without risking withering lawsuits from many of the largest tech giants on the planet.
.
ARM is buried under contractual poison pills. It has agreements with each of its licensees conceding the absolute right to any new chip designs created by the British firm. The is the reason Apple, Google, and the rest have never given the barest consideration to an ARM buyout.
.
If Bob is right and this is Softbank’s plan, it has the potential to be the single worst tech sector purchase of the past decade, and that’s saying something. It won’t be surprising if licensee lawsuits fly before the purchase is finalized.
.
Perhaps Softbank doesn’t realize the extent of these contracts? Perhaps Softbank naively believe AMD, Apple, Microsoft, Qualcomm, Samsung, and ever last one of the rest can be negotiated out of their rights to any new chip designs?
.
Big companies make mistakes. It’s not unknown for massive purchases to be based upon faulty beliefs. HP’s purchase of Autonomy and Microsoft’s of Nokia come to mind.
.
If Softbank truly believe they can steamroll a dozen of the largest tech firms on the planet into relinquishing their contractual rights to ARM’s designs, they’re in for a frightful surprise. Perhaps the only carrot likely to entice the licensees might be a relinquishment of all license fees going forward, but that would carry massive risk and destroy ARM’s existing revenue stream.
.
Were the license fees relinquished, it would be difficult to prevent the ARM licensees from starting their own version of ARM. A community owned chip design firm, working only for them, extending upon ARM’s own designs, while paying nothing to Softbank-ARM.
.
Even if Softbank were to negotiate their way out of their binding license agreements, there is no guarantee that Softbank’s ARM designs would win the Internet of Things battle. Samsung, Apple, AMD, or any of the others could take the lead, with products based on ARM’s own designs.
.
To describe the Qualcomming of ARM as a high risk scheme substantially understates the extent of the gambles involved, both on the legal front, and with the business case itself.
“Microsoft’s of Nokia come to mind. ”
That wasn’t a faulty purchase. That was a purchase to save face – in this case Microsoft’s and more specifically Ballmer’s. If Elop hadn’t been at the helm of Nokia pushing for Windows Phone it’d be a very different mobile world right now, perhaps with a lot of Android-based Nokia devices instead of the massive amount of Samsung devices.
Microsoft knew exactly what they were buying.
DISCLAIMER: THIS IS JUST MY OWN OPINION. IT’S VERY LIKELY FLAWED.
There can be a big gap between what Mr. Son would like to do and what will actually happen.
ARM Holding is not a simple box where you can put people in and from which you can take people away.
it’s a high tech company.
Well, all companies are made by real world men and women. It’s not the same whether the head of an office it’s me or you. Especially when you have been there for years with good results.
What Mr. Son wants to do is just a financial game, no matter what ARM Holding does: hamburgers or needles, air carriers or pencils. He thinks he has the money and he can do it. Nothing can be wronger than that!
ARM Holding has been doing well because of its human resources in first places and because of its politics and strategies in the second one.
By changing them, both of them, Son is replacing the company previously known as ARM Holding with another one with the same name but a rather different “soul”.
Mr. Son has no doubt: he’s smarter than the company board of directors, the company CxOs and the company directors. He’s sure he can do a miracle just because he thinks he can.
I can see three scenarios:
1. Mr. Son will succeed. My best wishes. To the company.
2. Mr. Son won’t succeed and will re-sell the company to China with a possible markup.
3. Mr. Son won’t succeed and will kill the company and will sell residual IP to China with a possible big loss.
In my (sick) mind, I don’t think option #1 will apply.
I appreciate the attempt to add insight, but you don’t know what you’re talking about just like with Apple’s investment into Didi. You would be better served asking more questions instead of assumptions.
(1) Qualcomm owns ATI mobile / aka Adreno GPU. It also owns various accessories like charging, capacitive sensing, and other bundled blocks.
(2) ARM is more like AMD, without the extra assets to design custom solutions for Microsoft or Dell.
(3) The problem is simple, as it stands without additional investment:
– Do nothing, market will expand into server designs and low-low-power IoT will grow.
– Raise license fees. Qualcomm, Samsung, Apple are initially hurt. They can quickly adapt and switch to MIPS or another core. ARM’s business is destroyed
Ultimately the CPU design is a dead business, just like Google and Facebook did to server design. Unit costs will be driven down to material & energy inputs. CPUs will be like RAM in the 90s.
Look up the RISC-V core.
Apple bought PA Semi in 2008, not ARM. This year it tried to convert its minority ownership in Imagination to buying it, for PowerVR graphics.
Intel bought Altera.
Microsoft is partnered with Xilinx.
NVidia has tapped into a vein of gold with machine learning.
The differentiation will be in graphics/parallel/dsp/FPGA acceleration, or in memory like Xpoint/ReRam/MRam. The frontier is doing more using less power.
There’s nothing special about ARM’s cores and ISA just like there’s nothing special about Intel’s fabs anymore. They surfed the wave until the beach, and one wave later everyone caught up.
You could make up a dream thesis about the loot to be made in modems, with the thirst for bandwidth, 5G, drone cell stations, micro cell stations, etc. It doesn’t matter how technically difficult, or useful to the end user… there isn’t big money in modem design, look at everyone that bought a modem company to help pitch their mobile CPUs. The telecom hardware companies are shrinking too, and look at how much more we use their hardware than ever before.
I’d like to see some grand plan where Sprint, Aldebaran, and ARM have some kind of magic synergy towards relevance tomorrow… but I don’t have a reason to expect that from Masayoshi Son’s ownership.
Türkiyenin en ideal araştırma firması olan emlak dream ile mecidiyeköy konut projeleri ayrıcalıkları ile 1+1 , 2+1, 3+1, 1+0 ve daha bir çok konutu
sizlere çok uygun fiyat seçenekleri ile sunuyor. Binlerce firma arasındaki verileri derleyip sizlere sunarak ev sahibi olabilmenin
avantajlarını sizlere göstermeye çalışmaktayız.
İstanbulun en işlek ve en gözde mekanında fevkalade konutlarda çok uygun fiyatlara sizlerle.
Ayrıca geleceğe yatırım alanında beylikdüzü konut projeleri çok değerli gayrimenkul değerindedir.
Yeni yapılar ve yeni yatırımlar ile güvenilir ve uygun konutları geleceğe yatırım olarak kullanabileceğiniz alanları sizler için
derledik incelemenizi tavsiye ederiz.
İyi Günler…
https://www.emlakdream.com/
Fascinating comparison between the different chip design companies valuations – strong support for your argument and the argument gets stronger each year. However, Britain’s biggest retail bank Lloyds has a tiny valuation relative to American banks and is efficient and very well capitalised so only having a US listing seems to increase relative valuations generally (Sterling has fallen 30% relative to Yen recently).
Is it really for the IoT potential as SoftBank claims ? I’d imagine the next three years could see ARM really start to hurt -all- the alternative architectures at all scales. ARM already dominates all the mobile device market. Intel makes wonderful server processors but it needed to withdraw from the cheap processor market as the latest ARM processors like Kirin 950 and Apple A9/10 chips completely overtook the Atom series leading Apple to claim that the A9 on a little battery was faster than 85% of Intel portable computers shipped last year. Intel is being squeezed by “Good enough” performance, battery capacity, heat generation and ever faster ARM. If ARM can also take much of the server market when Google and Amazon reveal their own chip designs, then all the other architectures will start to hurt and ARM dominance spreads wider and deeper. The fastest computers in the world are going to be European and Japanese ARM64 supercomputers – the specifications for both have already been published. Maybe then, when ARM dominates everything from phone to supercomputer, then the temptation both to design and sell their own chips would be too much to resist.
ARM has zero hope in servers. Literally no software, Xeon-D from the yesteryear wiping the floor with their best both in IPC and energy efficiency, which by itself is a horrible joke.
And considering the whole phone market reached the point of commodization two years ago, this purchase makes even less sense.
There’s not NO software: https://www.redhat.com/en/about/blog/long-arm-linux-red-hat-enterprise-linux-server-arm-development-preview with RHE and your favorite open source LAMP, apache or Hadoop compiled on it you have quite a lot of software available, even if you have to put some effort into it. Which people are doing. As for performance, there’s classes of operations for which more parallel cores can outperform fewer, faster cores. Those are the likely targets for ARM based hardware, not necessarily ALL server operations. As a disclaimer, I do work for ARM, do not speak for ARM, and I have nothing to do with cores – I know little past what I read in the trade journals.
Given that, however, I will also point out that I have heard nothing to the effect of what Bob is theorizing about the future direction of ARM under Softbank. We’ve only heard what everyone else has heard about the IOT angle.
But I will say, I think you’re all barking up the wrong IOT tree. ARM is well positioned to win in the hardware for IOT business. It probably doesn’t need Softbank’s money to sell microprocessors or other IP into IOT hardware. (Disclaimer, again, I don’t have anything to do with IOT, either, my opinions are based on what I just read what I read in the press, I’m not authorized to speak for the company, etc.) but read this carefullly: https://www.arm.com/markets/internet-of-things-iot.php and look carefully at this:
“The ARM® mbed™ IoT Device Platform is the fastest way to create commercial and interoperable IoT products based on ARM microcontrollers. The IoT is about devices and services in end to end solutions” emphasis mine. It’s that “End-to-end solutions” part that is the real lure of IOT.
Ok, let’s all set up an appointment on the calendar and come back here in 5 years to find out who’s right.
Extrapolation of the trends gives a pocket of predictability and a serious problem ahead for Intel. If all the microprocessor companies act in their own selfish short-term interests then ARM market share increases continuously just like it has been for the last 20 years. However, a big relative increase from 4% market share to 40% market share does little or nothing in terms of market control but the little increase from 40% market share to 80% changes everything. SoftBank needs to do nothing new and just let the current trends continue and let each ARM design licensee do all the commercial work for it to deliver an effective monopoly on all scales within a decade ! How good is that for a business plan – unbelievable !
All the hard work has been done already improving each ARM iteration when funding was tight and competition in mobile was fierce. Its only getting easier: UK salary costs just decreased thanks to Brexit, ARM has already won the mobile chip wars and is increasing rapidly on supercomputers and Internet of Things/Embedded microprocessor share. Intel has everything to lose and nothing to gain in terms of server microprocessor revenues so even the threat of competition will hurt its financial situation. It behooves each big Intel server customer to take an ARM license just to scare down Intel server profit margins. Its hard to imagine anything that can prevent a complete ARM takeover eventually ?
If ARM is so great then why do I despise using my one year old ARM based phone compared to my 7 year-old Intel based PC? They both have a 5″ screen and fit in my shirt pocket. Granted the PC doesn’t make phone calls, and would be unsuitable for that due to it’s 4 hour battery life, but for all computational chores, it’s a pleasure to use by comparison, and a modern PC desktop is even better. Microsoft embraced ARM with Windows RT, and Windows Phone, both of which were a disaster.
I personally blame the general uselessness of handhelds for anything but consuming content and awkwardly poking out text messages on Steve Jobs for removing styluses and keyboards from personal devices. If you have flat or pudgy fingers the miserable flat screen typing experience ruins everything. 🙁 I miss my stylus and graffiti!! (pocket PC had a great graffiti interface, too). You can get graffiti for android, but it doesn’t format the box the correct size due to all the different screen resolutions, and again, flat fingers. 🙁
Oh, that and the mobile versions of websites seem universally awful. This is probably to get you to switch to their app, which again I blame on Steve Jobs. Basically, everything you hate about computing on your handheld is not a computational issue, but an ergonomic one. If you hook a keyboard up to it and request desktop versions of sites, (like you can do with many convertible tablets) it is just as nice as a pc.
Perhaps it’s a chicken-egg thing. The ARM equivalent of existing Intel hardware won’t be put into desktop-like systems until the sophisticated software is ready to use it, but no one will write software for a device that doesn’t exist, so they only write ARM software for handhelds. So right now the only solution for me is to use an ARM device for 16-hour battery life on a pocketable device, like a phone, and full featured Windows on Intel for everything else. Microsoft is now pushing so-called UWP apps which can run on Windows 10 or Windows 10 Mobile, but so far all they’ve done is reinforce my belief that the old x86 programs on Intel are better, if only because, as you say, UWP apps have touch-screen handhelds in mind.
ARM and Softbank are aiming for 20% of the Intel server market by 2020 !
https://uk.finance.yahoo.com/news/chip-designer-arm-sees-20-060546016.html
I assume TMC is a typo for UMC?
While I’m here, I’d like to give a shout out to GLOBALFOUNDRIES. They own fabs in Singapore (former Chartered SC), Dresden (former AMD), and New York (former IBM plus a shiny new one in Malta).
In searching Google for those acronyms I came up with “Figure 4.4 Genealogy of Taiwanese semiconductor firms. Spin-offs included UMC, TSMC and TMC, … The stated intention of UMC to focus its activities on foundry work and to spin off its existing product-oriented … https://books.google.com/books?id=mJvWdaWc3Z0C&pg=PA193&dq=spin+off+TMC++TSMC+umc&hl=en&sa=X&ved=0ahUKEwi3k8aW8Y_OAhUD4GMKHfhCAZkQ6AEIHjAA#v=onepage&q=spin%20off%20TMC%20%20TSMC%20umc&f=false
its really simple ARM is the future of IOT. IOT is the future minirzation of cellular tech and low power communications.
Bob said “Five years from now buying ARM Holdings for $32 billion will look like it was a steal.” So if you agree with that, why are you saying “Cringelyisclueless”?
Got your IBM book. Fantastic read. Would not be buying shares despite Warren’s inclinations!
When Masayoshi Son, the billionaire Japanese technology investor, solidified his control over his SoftBank internet conglomerate last month, he told shareholders he still wanted to “work on a few more crazy ideas.”
https://taivideoyoutube.com/watch?v=NxfwRcJIICk
Hmmm…that domain is not youtube.com .