Season 4, Episode 2: ARM & SoftBank

[0] Because I mean, what's cool is like, it is literally the arm 610 that was developed with Apple for the Apple Newton is the core of all the arm processors.

[1] Welcome to Season 4, Episode 2 of Acquired, the podcast about technology acquisitions and IPOs.

[2] I'm Ben Gilbert.

[3] I'm David Rosenthal.

[4] And we are your hosts.

[5] Today, we are going to explore a topic that has flown relatively under the radar despite being the primary component of every single.

[6] one of our phones.

[7] SoftBank's $32 billion purchase of the British -based arm holdings.

[8] For folks that listened to the previous episode that we did on SoftBank, you know they were once a Japanese telecom and multinational conglomerate now that has a close to $100 billion fund, that they have created massive disruption in the startup landscape.

[9] And here we are diving into kind of the deal that started at all.

[10] with just an essential piece of technology that we all use every single day.

[11] Listeners, you know that a few months ago we started our limited partner program for folks to go deeper on technology, startups, and VC topics with us.

[12] David, I wanted to say that I'm particularly pumped where we landed in our last episode on investment theses on ambient computing and processing happening everywhere as sort of the next enormous technology wave.

[13] So I think it's super relevant to this episode and we will find out why as we dig in.

[14] So if you're interested or you just want to support the show and like what we do, you should click the link in the show notes to consider becoming a prestigious Acquired limited partner or go to Kimberlite .fm slash acquired.

[15] Okay, listeners, now is a great time to thank one of our big partners here at Acquired, ServiceNow.

[16] Yes, Service Now is the AI platform for business transformation, helping automate processes, improve service delivery, and increase efficiency.

[17] 85 % of the Fortune 500 runs on them, and they have quickly joined the Microsofts at the NVIDias as one of the most important enterprise technology vendors in the world.

[18] And just like them, ServiceNow has AI baked in everywhere in their platform.

[19] They're also a major partner of both Microsoft and NVIDIA.

[20] I was at NVIDIA's GTC earlier this year, and Jensen brought up ServiceNow and their partnership many times throughout the keynote.

[21] So why is ServiceNow so important to both Nvidia and Microsoft?

[22] companies we've explored deeply in the last year on the show.

[23] Well, AI in the real world is only as good as the bedrock platform it's built into.

[24] So whether you're looking for AI to supercharge developers in IT, empower and streamline customer service, or enable HR to deliver better employee experiences, service now is the platform that can make it possible.

[25] Interestingly, employees can not only get answers to their questions, but they're offered actions that they can take immediately.

[26] For example, smarter self -service for changing 401k contributions directly through AI -powered chat, or developers building apps faster with AI -powered code generation, or service agents that can use AI to notify you of a product that needs replacement before people even chat with you.

[27] With ServiceNow's platform, your business can put AI to work today.

[28] It's pretty incredible that ServiceNow built AI directly into their platform, so all the integration work to prepare for it that otherwise would have taken you years is already done.

[29] So if you want to learn more about the ServiceNow platform and how it can turbocharge the time to deploy AI for your business, go over to ServiceNow .com slash Acquired.

[30] And when you get in touch, just tell them Ben and David sent you.

[31] Thanks, ServiceNow.

[32] All right, David, I heard a rumor that you have one more thing for listeners.

[33] Indeed, one more thing.

[34] We passed a million downloads in the lifetime of acquired.

[35] Huge milestone.

[36] And to celebrate, we are going to do, you heard it here first, a worldwide virtual acquired meetup.

[37] So we're going to host this in Zoom.

[38] Ben and I are going to be on video and we'll have everybody else in the Slack community.

[39] We've got a bot installed that's going to do aggregate AMA questions and we're all going to hang out.

[40] So we're going to do it on February 21st, 2019, at 5.

[41] 30 p .m. Pacific time.

[42] We thought that would give everybody, you know, at least in the U .S., be 830 East Coast, 530 Pacific, and it's 9 .30 a .m. in China, for our listeners in China.

[43] So be there, be on Slack and on Zoom, and we will all hang out then.

[44] David, it's time.

[45] It's time.

[46] Indeed.

[47] Well, today we're going to be talking about arm holdings, and we're going to go literally from Isaac Newton to the Apple Newton to, to Masa and SoftBank and beyond.

[48] It's quite the scope, man. The first thing for listeners to know before we even dig in here, this chip company manufactures a total of zero chips, and in many cases they don't even design them either.

[49] Yeah.

[50] Well, we head to the prestigious and well -known Silicon Fen in Cambridge, England, home of Isaac Newton, inventor of modern physics at Cambridge University.

[51] the time is 1980 and in 1980 in England the BBC the British Broadcasting Corporation which controls the most popular radio and television stations across Britain and is a government entity they've teamed up with the UK Department of Industry and Department for Education and they're launching a major initiative that they're calling the BBC Computer Literacy Project and the goal is to educate the public and in particular young people at this time about computers and train them for the coming personal computer revolution that everybody thinks is right around the corner.

[52] And indeed it is.

[53] So they create a nationwide television series that they call the computer program.

[54] Get it?

[55] The computer program on television.

[56] Hey.

[57] That they show on, I think it was on BBC 2 to the whole country.

[58] It's a serialized program about how computers work.

[59] But the cornerstone and the really cool ambitious thing that they're doing is they want to put new micro computers true like personal computers for the first time into schools all around the country for children to learn and play on.

[60] So this is like super similar.

[61] So this is like a government version of Apple's strategy in the US at the time of getting Apple twos into schools so that kids can learn and play on computers with the idea that like eventually they'll grow up and then keep using Apple as they grow up.

[62] Google strategy today with the Chromebooks.

[63] Indeed.

[64] That I think is working really well.

[65] well.

[66] Yeah, I think so too.

[67] The BBC, together with the government, puts out a call for bids across all UK technology companies at the time to create this computer that they're going to put into schools.

[68] And they end up awarding the contract to a little company in Cambridge called Acorn Computers.

[69] Now, if you're a real computer history buff, you probably know about Acorn Computers.

[70] But who were they?

[71] So they started life in Cambridge.

[72] I believe they were initially called the Cambridge Processing Company or something like that.

[73] They made CPUs.

[74] It was a play on CPU.

[75] They made processors for other companies.

[76] But by this time, they'd started making their own fully integrated systems, you know, like the Apple 2.

[77] And supposedly, they chose the name Acorn because it sounded like Apple, but it was ahead of it in the telephone dictionary and directory.

[78] So when people were looking up computer suppliers in the telephone directory, they would see Acorn before Apple.

[79] Clever girl.

[80] Maybe that's how they won the bid.

[81] so they win the contract and they get to work on the device they're hard to work at it they know this is like a huge opportunity for the company in December of 1981 they start shipping it to schools around the country and it's branded as the BBC Micro and this is like a legendary computer in computing history like literally a whole generation of British kids grow up with this as their first, you know, exposure to computers.

[82] They end up selling over one and a half million units, almost all in the UK, which is like super good considering the Apple II was on sale for over 16 years and only sold six million units.

[83] So like, man, the power of like government sponsored programs.

[84] Crazy.

[85] So another thing, though, happened in 1981 that was pretty important, turned out to be far, far more important than the BBC computer program, which was that IBM introduced their first IBM PC in the American market.

[86] And the IBM PC was the first, like, personal computer that was, you know, because remember of computers before then, they were like in professional and business use.

[87] It was the whole client server model.

[88] You know, it was terminals, terminaling into the main frames.

[89] Mainframes, all that.

[90] like think about like what NASA was designing like the Apollo program on you know it wasn't personal computers so when IBM introduces the PC and it's targeted at business and professional users this is really like everybody in in the industry now is like okay wow like this is opening up this whole huge market where like it's the beginning of like software is eating the world right like now computers are going to be on every desk of every worker in every industry you know in the world at some point with IBM you know the power of IBM behind this so acorn realizes this and now like okay you know like the BBC micro is great this has been great for our company but like it's an eight bit micro computer like it's a personal computer but it's nowhere near powerful enough to compete with the IBM PC we need to we need to create our professional computer so they start a project they call it the Archimedes project this must be like a legacy of, you know, Cambridge and academia and, like, having these, you know, a scientist and mathematician names.

[91] They realized, though, the processor that they'd been using in the BBC micro, they hadn't designed it.

[92] It was an off -the -shelf 8 -bit processor from a firm called Moss Technology.

[93] It was called the 6502.

[94] This is when a processor had really cool names, like 6502.

[95] Yeah, super, super cool.

[96] I miss that these days.

[97] Yeah, I know, I know.

[98] Although arms naming's games are not really much better these days.

[99] Those don't end up becoming consumer brands the way that like, you know, A11X or something like that does.

[100] Although A11X, we've completely gone back, never mind.

[101] My point is completely destroyed by the fact that Apple is calling chips, like, it made sense when it was like A5, A6.

[102] Bionic was a couple.

[103] So Bionic actually means nothing.

[104] Yeah, well, I don't know, but they actually don't mean anything.

[105] It's a marketing brand that doesn't, it's not a modifier on the name of the chip, which is like A10 or A11.

[106] It is just a, It's kind of like, you know, MacOS 10 .4 Leopard, or I think that's wrong, but, yeah, anyway, too many digressions.

[107] Too many digressions.

[108] Anyway, so Acorn hears that Intel, off in California, you know, their processor company, everybody knows about at this point, they are working on the perfect processor that they can use in their new Archimedes project.

[109] It's going to be a 16 -bit processor.

[110] it's getting a ton of buzz and it is this is like a super famous in computer history this is the 80286 processor which would eventually become just the 286 which successors would be the 386 the 486 and then the Pentiums and then every modern Intel chip that we know today this was the first the X86 architecture if you will indeed this is the first of the X86 so acorns like oh cool like hey Intel like we want to be a customer to you guys, can you like, you know, ship us over some so we can build some reference, you know, PCs and start working with you guys.

[111] And for some reason, and this is going to go down as like one of, you know, there are a few of these moments unacquired of like, you know, history turns on like a knife point, you know, whether it's Blackbuster or, or here, Intel.

[112] Intel makes one of the worst business decisions in history, in all of business history and says, no. You guys, you know, BBC Micro, come on.

[113] You're some like little PC, you know, company in Cambridge, you know, England, not even Cambridge, Massachusetts.

[114] It's like, we don't care about you guys.

[115] Yeah, surely nobody will ever, you know, decide to do something on their own when we tell them no and then eventually lead to our own demise.

[116] I mean, we're Intel.

[117] Like, who else is going to make, you know, high performance computing microprocessors in the world except us, Intel?

[118] I noticed you said high performance.

[119] So I think that probably is still true.

[120] well it depends how you define high performance um so acorn they are they're kind of out of options and they're like well you know what can we do what do we have the one thing they have that basically you know very few other computer companies in the world even really apple at this point have they have a ton of super super smart physicists and engineers and and early in the burgeoning field of computer scientists, students from Cambridge University, where they're located, that they've employed either as interns or full -time staff.

[121] And these people are like super talented.

[122] And they decide, you know what?

[123] Maybe we can just build our own microprocessor.

[124] Like we started as a microprocessor firm in another era.

[125] Let's try it again.

[126] They decide, they give the task to a team of engineers on staff.

[127] And they say, okay, we need something super high power.

[128] that can essentially compete with this, with Intel's 286 chip, we need a, we need almost that good level of performance, but it has to be super low cost because like, we're making this ourselves, you know, we don't have the global resources of, of Intel behind us.

[129] What can we do?

[130] And so this team led by Sophie Wilson, who was and is an incredible computer scientist, you know, one of the most important, you know, contributions to all of computing as we see that she implements here.

[131] She'd heard about a paper out of, ironically, back in the Bay Area in California, out of UC Berkeley that has just come out, about detailing what a project would look like for a reduced instruction set computer.

[132] Now, this is going to be a technical interlude, but it's super important to understanding, you know, again, what becomes probably one of if not the most important underlying technology company for the whole industry right now.

[133] So what are instruction sets?

[134] Ben, do you want to talk about this?

[135] I will.

[136] And I first off, want to take a step back here and make a disclaimer that we acknowledge that we are dramatically oversimplifying technical details in this episode.

[137] We do this with apologies to those who find this too basic, but also conversely, for those who find it to be esoteric on the other side of the spectrum, we are going to dust off the computer science degrees here and died in.

[138] I would say in true Apple fashion here, this show is, and really the whole technology industry is the end.

[139] intersection of, you know, engineering and the liberal technology and the liberal arts.

[140] And that's what this is here.

[141] You're about to get a dose of technology.

[142] May we be so bold.

[143] So before we dive into what is a reduced instruction set architecture, it's worth diving in what came before it, the complex instruction set architecture.

[144] And even before that, like, what the heck is an instruction set architecture?

[145] Like, why are we, why?

[146] Okay.

[147] So the instruction set architecture, you can think about as sort of the language of a chip.

[148] So, a CPU has a variety of components on it, places where you can store information, places that temporarily hold information, places that tick the clock and move all the information one step forward in the process.

[149] And the way that it does all of this saying, hey, go store that over there and that register.

[150] Or, hey, advance the whole clock, you know, one step so that we can move this thing out of that register and push it off into memory or something like that.

[151] All of this happens.

[152] And this is all what's happening inside.

[153] the CPU in your computer, whatever device you're using right now.

[154] Yeah, and at what speed?

[155] I don't know, like hundreds of thousands of times per second or millions of times per second.

[156] I mean, this happens incredibly fast, but it requires a language.

[157] It's sort of like its own programming language, and it's the instructions or the instruction set that the chip itself speaks.

[158] And so originally, the way that this was done was with SISC.

[159] And this was really the complex instruction set computer where instructions were, well, first of all, there were a lot of them.

[160] They were very, very sort of malleable.

[161] So they could do things like take multiple clock cycles to achieve a complex instruction.

[162] And this is like multiplying numbers, dividing, transforming numbers, all sorts of stuff.

[163] Right, right.

[164] It would sort of do as much as it possibly could using the hardware circuitry.

[165] So it was really about, hey, this is going to be powerful hardware, so we're going to write a language that leverages all the very unique and powerful components that are actually on this chip.

[166] It could have, without getting too complex on these things, variable length instructions.

[167] So you couldn't rely on a set of assumptions about how long each instruction was going to take time -wise or was going to take up space -wise.

[168] So while it provided the programmer with a lot of power, it required a tremendous amount of sort of tight coupling with the hardware and complexity.

[169] Yeah, and that translated to, at the time, everybody was thinking about number of transistors on the chip and associated memory that you needed to support the data that was going through all those transistors over time, and much more importantly, that would translate into power consumption.

[170] The SISC, CISC complex instruction set architecture computers, that is what Intel processors are.

[171] So Intel, like, the set of instructions that is SISC, comes from Intel, and all Intel X -86 processors are CISC, complex instructions that computers, what ACORN at this time, and Sophie Wilson, what they were picking up on out of academia was like, maybe there's a different approach to how you could build these processors.

[172] I've actually never seen this written about as low -end disruption, but this is almost like classic low -end disruption where we say, you know, what if we punt on a lot of those things that everyone previously thought was important and solve a problem with a very different set of constraints where it's got to be cheaper.

[173] It's going to be less sophisticated, so there's less components on the chips.

[174] There's going to be less instructions that are available for the programmer to use.

[175] And really, it's not the programmer.

[176] It's actually the compiler that translates what the programmer writes into the instruction set.

[177] But it's really rethinking it from the ground up and saying, what if every instruction could only take one clock cycle?

[178] It could work on sort of a variety of different components, but making this really hardcore set of assumptions.

[179] It was this software -centric design instead of a hardware -centric design, so it was kind of portable.

[180] It could use a very limited number of addresses and limited number of registers.

[181] So in all aspects of it, it's sort of worse.

[182] You could sort of describe it as worse in every way.

[183] But I'll turn it back to you David for so how does that actually end up being disruptive and better what one of the team member original team members who worked with sophie on designing this would write later you know you could think of it as it's the 8020 rule right like these risk reduced instructions set architecture the machines processors that that acorn designed they were they only could do about 80 % of the instructions that sisk could do but they did those instructions they executed them much, much faster, like on one clock cycle instead of multiple clock cycles per instruction.

[184] So when you needed to do those complex, you know, other instructions that were in the 20 % the chips would slow down hugely.

[185] But the thing was you just didn't do them that much.

[186] So like the 80 % that you were doing most of the time, like they were really, really good at.

[187] You were spending a lot of money to build this sort of sophisticated components onto the chips so they could handle those things that you just weren't doing that often.

[188] To put some numbers on.

[189] this.

[190] The 2286, the famous Intel processor that started the x86 line, the first one of those had about 130 ,000, 134 ,000 transistors on the chip.

[191] The acorn processor, the risk processor, when they finish it, has only about 30 ,000 transistors on the chip.

[192] And yet, because of this 8020 rule, it actually has better performance than the 2286.

[193] So it's like way cheaper to make, and it has actually better performance for most applications.

[194] This is like huge.

[195] This is a huge leap forward in engineering and computer science.

[196] So they do this.

[197] Sophie and the team create this like in a very short period of time.

[198] They decide that they're going to call this chip the acorn reduced instruction set computer machine.

[199] A .R .M. Arm.

[200] Boom.

[201] And the world changes.

[202] But not just yet.

[203] The groundwork was was laid for the world to change, but it wasn't yet.

[204] It's not like, you know, here we are in the 80s and Intel starts doing poorly.

[205] You know, these things take a long time.

[206] This is really interesting.

[207] So now let's bring, we've just went deep technically on what's going on.

[208] There's this huge innovation.

[209] Why didn't the first arm chip and the Archimedes that launched in 1987 with this chip, why didn't it blow the IBM PC and the IBM PC clones out of the water?

[210] Well, this is like a market forces thing.

[211] so by the time it launched Microsoft was on the scene right and DOS was around and DOS ran on Intel X86 architecture and that eventually Windows would run on top of DOS and started getting so much market share and all the applications that all these business and professional users needed they were all running on DOS DOS didn't run on Archimedes or the risk architecture computers so for a long time for the whole entire PC wave everybody in the industry knew that other types of architecture, risk -based architecture, which aren't pioneered, but the Motorola Power PC, like what Apple was using at the time, it was fundamentally better technology, but because of the duopoly between, the Wintel duopoly between Microsoft and Intel, it's almost like nobody really cared.

[212] It's important to understand what role a compiler plays in all this.

[213] So why is it that they couldn't just run the run windows, run DOS on, you know, these better chips.

[214] Until recently, and there's been a lot of really great advancements in really commercializing this sort of crossover technology recently, it was thought to be basically impossible to make your programming language, which is, you know, written in C and using sort of the standard C compiler, work well across different chip sets, especially when, you know, so many layers of translation from the application to the operating system, to the kernel, you know, eventually getting, you know, actually executed on a chip.

[215] And there's a lot of sort of assumptions that are baked into the programming language, to the compiler, to the chip, and sort of assumptions around those things being coupled together.

[216] And so it's sort of much more difficult than you would think to rip the layers apart and say, well, we're just going to run this on a completely different architecture.

[217] You know, Sophie and team did this amazing thing of coming up with for a, you know, for a different scenario, a much better instruction set architecture.

[218] However, basically nothing would run on it and they needed to kind of think about the world in a whole different way and convince everyone else that they should think about the world in a whole different way in order to leverage that innovation they created.

[219] Which wasn't going to happen during the PC wave because, you know, the network effect flywheel was like in full swing at this point, you know, you've got Lotus, you've got, you know, all these application providers writing for, as we talked about, writing for Microsoft, which only worked on Intel.

[220] So boom, there you go to, you know, multi -a -hundred billion -dollar companies come out of that wave and Acorn is left in the dust.

[221] So as you would think, they're kind of in a bleak position here now.

[222] They've put a ton of resources into this new Archimedes project that they think is going to be, You know, they're going to ride this next PC wave.

[223] They're getting knocked off the wave.

[224] You know, this is probably the end of the company, right?

[225] And actually, it is.

[226] Acorn itself ends up getting acquired by an Italian computer company called Olivetti around this time.

[227] And they end up just exiting the PC business altogether.

[228] But there was something also that we haven't talked about yet that was pretty interesting about these arm processors that they were building.

[229] and this is another just like crazy thing of history that like nobody they didn't expect it so when they developed the arm processors when sophie and the team developed it the goal was same level of performance ended up being slightly better for a lot less cost like put fewer transistors on the chip because that's going to cost less but what it turned out they would put these these processors into the PCs into the archimedes prototypes it turned out they sometimes would function even without a power supply and the team at first they were like this is like crazy what is they like a ghost in this machine like how is this how is this processor working without a power supply was it like what residual well it was it was without a dedicated power supply to the processor and this is like crazy because at the time like you know CPUs were the most power hungry components of the whole you know PC so like there was power running into the machine, but just not a dedicated line into the CPU.

[230] And it turned out that because there were so many fewer transistors on the chip, it needed much less electricity and power to be able to run, even at really high performance.

[231] And so it was actually just sucking power from the other components in the circuit on the chip.

[232] And it was still able to function.

[233] And they were like, whoa, this is crazy.

[234] So it turned out after Acorn gets acquired by Olivetti, one of the two original co -founder, this guy, Herman Houser, who also was a Cambridge physics PhD, he leaves, and he's thinking, which a couple of people are thinking in technology at the time, you know, PCs are here, but what's the next wave going to be?

[235] This is like, you know, people thinking about VR right now or a couple of years ago.

[236] Like, we're in the middle of the mobile wave, but like, what's next?

[237] And it's worth sort of like simplifying what we have right now.

[238] We have a thing that uses a completely different architecture that's not really compatible with everything the rest of the world uses, but is much cheaper to make high performance and requires less power.

[239] And so Herman's like, you know, I think mobile computing, like everybody's all about desktop PC computing right now, but like think about everything you could do if you had a computer around in the world with you, like not tethered to a desktop.

[240] He goes and he starts a company to try and pursue this vision of computing.

[241] He calls it the active book company.

[242] And he wants to focus on making PDAs, personal digital assistance.

[243] which are mobile devices.

[244] And what's super important for mobile devices?

[245] Battery life.

[246] And this is 1988?

[247] This is 1988.

[248] So battery life is very important.

[249] Right.

[250] Very important.

[251] I think lithium ion batteries get like 15 % better every year.

[252] So compound that back 30 years.

[253] Like they were bad.

[254] Yeah, right.

[255] Like how long does your iPhone last today?

[256] Like not long enough.

[257] You know, imagine back then.

[258] but there are just not that many people in the world that are thinking like 1988 shoot like I think I got my first computer at that point I was like four years old and it was a massive like you know hunk of metal that sat on my desk and probably used you know half of the house's power supply like well was an x86 what do you expect yeah right exactly so so Herman he's like he's focused on PDAs and he's like you know what I think could make this happen is these processors we developed back at arm So he goes back to the chip team at Arm and to Sophie.

[259] And he's like, hey, can you like rework this whole thing to really optimize for power consumption while still maintaining this high performance?

[260] So he's like, yeah, we can do that.

[261] And they do.

[262] And it works.

[263] And they start producing these chips with their longtime silicon partner, VLSI.

[264] because, again, ACORN didn't have the resources to be manufacturing their own semiconductors at this point.

[265] So they used a fabricator, VLSI, to make their silicon.

[266] Well, at the same time, who else is thinking about PDAs out there?

[267] This is 1988, Apple.

[268] And this is where the other Newton, not Isaac, comes into play.

[269] So Apple, Apple's under, Steve Jobs is left at this point.

[270] He's getting kicked out of the company.

[271] John Scully is CEO and John Scully's great white whale is the same thing as Herman Houser it's the personal digital assistant you know Scully gets like such a bad rap but like he actually had you know a vision that like very few people in computing you know did at the time not even Steve I mean Steve was off doing next right like he's trying to make a super powerful workstation come on what's the difference between being wrong and being too early nothing but But good for the world that he was wrong and too early because Scully puts Larry Tesla at Apple on this Newton project, which is his pet project.

[272] And Larry is amazing.

[273] So Larry came from Xerox Park.

[274] He was like an OG original computer guy.

[275] He invented copy paste.

[276] He's like leading the SWAT team within Apple that's going to build this, you know, computing platform of the future.

[277] He knows he needs a low -power, high -performance chip.

[278] He first goes to AT &T, which was working on a low -power chip called The Hobbit, which you can't make this stuff up, just like you would expect from a chip from AT &T called The Hobbit in 1988, it sucked.

[279] Like, it was terrible on every dimension.

[280] And this is fun.

[281] I think one of my carve -outs a few episodes ago was Jerry Kaplan's book, Startup.

[282] and Jerry, he was the founder of the Go Corporation here in Silicon Valley, which was also trying to work on a PDA at the time.

[283] They had ended up getting acquired into AT &T and formed the backbone of The Hobbit.

[284] Anyway, Larry gets introduced, he's chatting with people in the Silicon industry, trying to find better chips.

[285] He's chatting one day with someone at VLSI, and they're like, hey, you know, we've got this partner, Acorn, over in the UK, and like, they're not doing too well.

[286] you know, they were trying to compete with you guys in the telephone directory, but one of their founders is doing something kind of interesting right now, also in mobile computing.

[287] And they have this chip that's working pretty well.

[288] You might want to check it out.

[289] And Larry's like, I need to have this right now.

[290] So he goes over.

[291] He meets with, uh, with Sophie and the engineering team over there.

[292] And he's like, this is great.

[293] This is the chip we need here at Apple for the Newton.

[294] but I can't license this tech from Acorn and Olivetti like we're competitors we need to we need to create a structure that can work here so they architect a deal with Oliveetti and VLSI they spin off the chip division of Acorn into a new separate company that they can then license this processor design from they do it super fast within like six weeks of when when they get together they've spun out the company into a new division, a new, new totally separate company called Arm.

[295] This blew my freaking mind that Arm was started as a JV with Apple on the Newton.

[296] It's crazy.

[297] So Apple invests $1 .5 million into the company.

[298] So Acorn and Olivetti are bringing the engineering group.

[299] They assigned 12 engineers to the new company.

[300] VLSI is the Silicon Fabricator partner.

[301] They get an equity stake.

[302] Apple, brings the money, they get 43 % of the company for one and a half million dollars.

[303] Again, talk about a dilutive seed round.

[304] To blow out the irony of this, I'm going to jump forward just to foreshadow like all of the A series chips are arm processors.

[305] Like the thing that is so differentiating about the iPhone, I mean, there's a lot of things.

[306] But one of the things that they're just years ahead on is being able to have some of the best processes in the world, all based on the arm instruction set architecture.

[307] We will get to how these companies have sort of parted ways over time, but oh my God, they were actually involved in the founding of it.

[308] It is no overstatement at all to say that without the Newton project within Apple, arm as a company would not exist.

[309] The phones, whatever device you're listening to this on right now would look very different.

[310] You might have it plugged into a wall, which you almost assuredly don't.

[311] It's crazy.

[312] It's crazy.

[313] So, these 12 engineers, they do the spinoff, they get the $1 .5 million from Apple.

[314] These 12 engineers go down the street in Cambridge.

[315] They set up shop in a converted barn and they bring in this guy, Robin Saxby, who had been an executive at Motorola to be the CEO of the new company.

[316] And they get to work with Apple.

[317] They're working super closely.

[318] They're taking the core risk processor technology that they own, but it needs to be super customized and fit into a chip set that's going to work with in this small device, the Newton.

[319] So they're working hand in hand with Apple to create it.

[320] And they make the processor.

[321] It's the Arm 610.

[322] And that goes into the Newton.

[323] And every Newton, the first Newton's that shipped, all had the Arm 610 processor in them.

[324] And then it would get upgraded over time.

[325] But it would power every Newton.

[326] Now, ironically, remember Herman Hauser, the original Acorn co -founder, who had started Active Book and kind of set all these wheels in motion, he ends up selling Active Book to AT &T of all people in 1991 and AT &T demands merges it in with the assets of go that it had acquired, which is now EO within AT &T, only they could do something like this.

[327] They force it to start using the Hobbit processor.

[328] And of course, I don't know, I think these PDAs did end up shipping, but like they were so bad and it completely fails in the market.

[329] David, the rabbit holes that you went down for this episode, I believe achieve new heights.

[330] Oh, man. This is so much fun.

[331] Speaking of failure, though, I mean, AT &T was, like, obviously stupid with, you know, the Hobbit processor.

[332] And go read the book startup.

[333] Like, it's so good about all the ridiculousness that went on around this.

[334] The Newton also, of course, famously fails, even though they have the arm processor.

[335] And it's just too early.

[336] Like, the world isn't ready for this.

[337] So catch me up.

[338] Like, I didn't research this part at all.

[339] How does, what happens to arm post -Newton failure?

[340] year.

[341] So when the Newton actually gets around to shipping in 1993, it's now been over two years that arms working exclusively with Apple.

[342] The Newton ships, and like it's clear pretty early, like there aren't going to be enough unit sales here that like the project's going to be powerful.

[343] It's like the home pod of PDAs.

[344] Yeah, it's the home pot of PDAs.

[345] So this is where Robin, the CEO from Motorola who came in, this is where he makes a couple really, really brilliant decisions and we've talked about in previous episodes that like one of my tech themes has been recently that like when you can marry a huge technology wave with a key business model innovation like we talked about in the 10th episode like that's when something magical happens and so what did robin do he was like okay well we need to work with other partners here we can't be dependent on just apple and just newton to buy a lot of chips from us because they're not going to they're not going to pay us a lot in licensing fees here but he's also realized that like part of the magic of what made this processor really work wasn't just that like arm designed this chip and like handed it over a whole cloth to apple it was that they really embedded with the engineering teams on the Newton and made something like pretty customized that worked specifically for this device and so he's like you know I bet we could do this with lots of people and we give them this core processor, this core risk processor technology, we could embed within their teams and we can help them develop essentially custom silicon for their use cases.

[346] And you know what is really interesting here?

[347] Like if we could create a business model that aligns with this, what if we say like, okay, when we do this, you pay us an upfront licensing fee for the rights to our core risk technology and you pay us for our engineering time for embedding with you and we'll make money on that but let's align get aligned on like actually shipping units we want you to ship a bunch of units we want you to ship a bunch of units so how about we take a small royalty on every device you ship that has our technology in it and then to the time like this is like okay somewhat interesting like how many devices could potentially ship that are like mobile computing devices so arm gets gets paid three times like they get the license fee from hey you know you get to use the arm technology you know our this instruction set architecture is one that you have the privilege of using on your chip and there's sort of two ways that that can work either they design the chip for that that manufacturer or they say here you feel like you're a good designer and this is sort of how apple's relationship works today you use our instruction set and you do it so then they get paid the second time for actually embedding with them.

[348] That's their sort of software and services line of business.

[349] And then they get that third time, which is every unit sold, you know, they get a tiny little piece of the cost of each CPU that shipped out.

[350] Yeah.

[351] Well, you know, it turns out, I mean, this is one of the things that makes the technology industry magical.

[352] A very, very tiny piece of a pie that is literally almost 100 times bigger than every person in the world combined, which the number of devices in the world is at this point, turns out to be a very, very large slice of pie.

[353] Do you know how many arm chips have shipped to date?

[354] I do, but why don't you go for it?

[355] 130 billion.

[356] Yeah, that's billion with a B. And that's a lot of pennies.

[357] And think about that.

[358] Those are arm chips, each of which is the core processor in a device.

[359] So that is a hundred, 130 billion devices.

[360] Not necessarily true.

[361] Lots of devices have...

[362] Well, many devices have multiple chips within them.

[363] Yes.

[364] But still, like, you know, it's a...

[365] Okay, reduce that by, you know, a factor of four or five or whatever.

[366] Like, that's still, like, way more than there are people on Earth.

[367] Yeah, that was a thoughtful first contract to sign that they get some upside on units shipped.

[368] Yeah.

[369] And what's also cool.

[370] So, again, like, I think this is such a cool example, arm is, of...

[371] technology and business model playing together and making each other better.

[372] So, like, obviously there's the financial aspect of this.

[373] What's interesting on the technology side, too, because Arm now, like, all the other chip companies take Intel, for example, they're like, use our, you know, Adam 7, whatever chip or core eye, blah, blah, blah.

[374] I am going to give you this chip.

[375] You are going to put it in your device.

[376] Because Arm is like, no, no, no, we're aligned with you.

[377] We want you to make the best products and ship the best.

[378] devices and we'll embed our technology and our teams with you, we can collaborate on designing it however you want.

[379] So this is what really, it's arm and this model that starts to enable systems on a chip to be really take off.

[380] So what is the system on a chip?

[381] Like back in the PC days that we were talking about earlier, you know, a system, a PC, you had a motherboard, you had a processor, you had a graphics card, you had a sound card, you had a, I remember plugging all this stuff in and building my own PCs back in the day.

[382] Yeah, man. Put it all in your PCIE slots and call it a day.

[383] Exactly.

[384] Oh, man. Incredible.

[385] Is that PCI Express, right?

[386] Yeah, that was the latter generation.

[387] Yeah, that was like, oh man, crazy.

[388] So people start to realize, especially in a mobile environment, like, well, what if you could just put all of that just on one chip instead of having separate chips and buses and motherboards and whatnot?

[389] Arms like, yeah, cool.

[390] Like put our processor in one chip as part of it.

[391] but like we'll we'll help you with that and so apple now with the core you know the a a 57 you know bionic superman chip um that's what this is uh all the samsung chips qualcomm chips ti chips like every you know chip your phones have multiple chips but like at the core most of the technology is being done on one actual piece of silicon this was the transition from the era of sort of the just discrete CPU to two system on a chip which is you know, where we are today.

[392] Yep.

[393] And so actually that same year in 1993, and like, again, so much kudos to arm the company and to, to Robin the CEO for turning this around, the same year that the Newton ships and they realize this isn't going to work, they sign a landmark deal with Texas instruments to provide the core of a processor that Nokia has contracted with TI to go into the Nokia 610, or 6110.

[394] We'll try and put a link to this in the show notes.

[395] Once you see this phone.

[396] You were going to remember this phone.

[397] This is like, I think the first cell phone I had was like a variant on this.

[398] Yeah, this is the candy bar phone.

[399] This is the first major consumer GSM phone that is sold certainly in America and all over the world.

[400] And this is what starts, you know, the kink in the curve of cell phone shipments that ends up, you know, with smartphones and where we are today.

[401] David was talking earlier about why, you know, these arm chips were, you know, even though they were better in a lot of ways.

[402] They couldn't penetrate the duopoly of Microsoft and Intel and Windows running there.

[403] You can start to see now, you know, the software that ran on that Nokia phone that I'm sure everyone played Snake on is nothing like the software that ran on Max or on Windows.

[404] Snake was like the killer app because because armed processors were literally like before the Nokia 6110, the processors in cell phones weren't good enough to even run snake.

[405] But now you have this low power, high performance processor that can run games, right?

[406] And you can like draw a direct line from there to, you know, Farmville to clash, you know, clash of clans to everything.

[407] Right.

[408] The bridge that hasn't gotten crossed yet is how we went from that, which really, I mean, that phone felt more like an embedded device than really like phones as we know today.

[409] And phones today are effectively PCs.

[410] And if you think about the work that was done to create the first version of iOS.

[411] It was really to strip down macOS.

[412] And people always harp a lot on this like strip down macOS to create iOS and it still uses the same Darwin kernel and all that stuff.

[413] But the other crazy piece of work that had to be done to bring a computer operating system to these mobile phones was adapting it for the armed chip set.

[414] Like to bring PC operating systems to something that would operate on these chips that like weren't getting a lot of power and had you know, completely different instruction set.

[415] The fact that iOS and Android works the way it does today on this chip that was nothing like what they were originally architected for is mind blowing.

[416] You know, it's interesting.

[417] I didn't put two and two together till right now, but I strongly suspect, you know, if you think about, like, why did Microsoft miss mobile?

[418] There are lots of reasons, right?

[419] Like cultural, you know, what -have -you technology, you know.

[420] But actually, like this specific technology reason I think could be.

[421] could be one of the major points that, you know, Microsoft had Windows mobile, right?

[422] And Microsoft had Windows C and embedded Windows and whatnot, right?

[423] But, like, those were...

[424] Completely different code base.

[425] Completely different code base, right?

[426] That wasn't Microsoft Windows because Microsoft Windows and DOS only ran on complex instruction set architectures, right?

[427] They're not going to re -architect that whole thing to run on risk architecture.

[428] Now...

[429] I mean, they do, they do now.

[430] Yeah, they do now.

[431] But last couple of years.

[432] At the, you know, as smartphones were taking on, why could Apple do this?

[433] Apple had always been much more open about their, you know, about their architectures that Mac OS ran on, that OS10 ran on, right?

[434] And like, oh, did they already do the transition from the Motorola processors, the Motorola power PCs to, well, they had, yeah, they had to pour it into the X -86 Intel with OS10, right?

[435] So they were much less wedded to, you know, and open -deporting their MacOS and OS -10 into different architectures.

[436] And indeed, they had this Newton DNA as well.

[437] For folks who don't know specifically what we're talking about, so the iPhone came out in 2007, which, holy God, they put MacOS on an arm chip.

[438] Two years before that, the thing that we're referencing here is in a very Jobsian keynote that continues to go down in history as just an amazing piece of showman.

[439] ship.

[440] Steve Jobs came out and said, uh, so we're changing the, uh, chips that are in all the new Macs away from the power PC that we've been using to using Intel, which of course was so dramatized.

[441] And Paul Adelini walked out on stage and oh my God, you know, Apple is using Intel.

[442] And, you know, look at all the old wars that are there's still, the ground is burning from all the carnage and wreckage of those old wars.

[443] And here's what's happening now.

[444] And Steve, of course, says and the crazy thing is that you guys don't even realize it's been that way for a year and all the operating systems dating back, you know, a year or two years ago.

[445] I didn't know it was multiple years, yeah.

[446] Yeah, have been capable of doing this and you guys just didn't know it.

[447] So go buy yourself a new computer and all your old stuff will work on it.

[448] David, you're right.

[449] I think that that did give them sort of the confidence to say like, well, you know, I guess we could start re -architecting our operating system to work on yet another chip set.

[450] Well, and, of course, at that point in time, the iPhone project was well underway, probably getting close to shipping, because I think it was 2006, right, when that happened.

[451] I wonder how similar those efforts were.

[452] Yeah.

[453] Interesting.

[454] Well, anyway, to rewind back to Arm.

[455] So this is happening.

[456] They do this T -I deal.

[457] The Nokia 6110 launches.

[458] Things are all basically up into the rate indefinitely from then for Arm.

[459] So at the end of 1997, Arm is doing over 25 million pounds in revenue.

[460] they're profitable.

[461] They do a dual IPO both on the London Stock Exchange and the NASDAQ, because remember they have British shareholders, Italian shareholders, and an Apple, American shareholders.

[462] On April 17, 1998, they priced the IPO at 5 pounds and 75 pence per share, which translates to a market cap of 264 million pounds.

[463] Now, the exchange rate was stronger back then, but still, my God, I wish I could have invested at the IPO.

[464] It grows, hugely throughout the tech bubble, especially.

[465] And this is, this is so awesome.

[466] I can't believe this is going to be like a footnote in this episode, but like, I want to highlight here, this saves Apple.

[467] The ARM IPO had it not happened, very likely Apple would have gone bankrupt because this is 1998.

[468] Steve Jobs had just come back into Apple.

[469] Scully was ousted.

[470] The company's bleeding cash.

[471] Like, they are like seriously facing bankruptcy.

[472] And what?

[473] It's a decade later, but they still own that share of arm?

[474] No, no, no, no. They start liquidating the arm shares.

[475] And thankfully, they don't liquidate all at the IPO.

[476] They start selling slowly over a couple of years as the arm stock price is going crazy.

[477] Apple, remember, they invested one and a half million.

[478] They make $792 million in profit from selling their arm stock over the next couple years.

[479] And literally, that is what saves the company.

[480] Like, it would have gone bankrupt without that.

[481] Imagine being an analyst.

[482] Like now there's an army of Apple analysts, but like then, you know, imagine being an Apple analyst and you're like, okay, yet another quarter where all of their, it wouldn't even be operating income, but all of their profit is coming from liquidating this thing that's going to run out.

[483] I mean, that was literally what kept the company afloat.

[484] Kind of amazing while Steve was, you know, preparing what would become the IMAC and consolidating the product lines.

[485] And ironically, killing the Newton.

[486] So when the tech bubble bursts, when the internet bubble bursts and the telecom bubble bursts in 2001, Arm, of course, you know, takes a blip and shipments basically like plateau for, you know, a year, maybe not even.

[487] But again, like this is such a huge wave.

[488] Like by 2002, they pick back up to the point where in 2010, Arm is now, arm partners are now shipping over five billion devices every year.

[489] again, so almost as many people as there are on the planet, arms partners are shipping devices.

[490] So like, you know, these are, these aren't just, of course, cell phones.

[491] These are, you know, microwaves.

[492] These are refrigerators.

[493] These are cars.

[494] Like cars have a lot of arm chips in them.

[495] These are sensors and devices, anything that needs a low power, high performance embedded processor.

[496] All right, listeners, our sponsor is one of our favorite companies, Vanta, and we have something very new from them to share.

[497] Of course, you know Vanta enables companies to generate more revenue by getting their compliance certifications.

[498] That's SOC2, ISO 2701, but the thing that we want to share now is Vanta has grown to become the best security compliance platform as you hit hypergrowth and scale into a larger enterprise.

[499] It's kind of wild.

[500] When we first started working with Vanta and met Christina, my gosh, they had like a couple hundred customers, maybe.

[501] Now they've got 5 ,000, some of the largest companies, out there, it's awesome.

[502] Yeah, and they offer a tremendous amount of customization now for more complex security needs.

[503] So if you're a larger company, and in the past, you showed Vanta to your compliance department, you might have heard something like, oh, well, we've already got a compliance process in place, and we can't integrate this new thing.

[504] But now, even if you already have a SOC2, Vanta makes maintaining your compliance even more efficient and robust.

[505] They launched vendor risk management.

[506] This allows your company to quickly understand the security posture of the vendors that you're choosing in a standardized way that cuts down on security review times.

[507] This is great.

[508] And then on the customization front, they now also enable custom frameworks built around your controls and policies.

[509] Of course, that's in addition to the fact that with Vanta, you don't just become compliant once, you stay compliant with real -time data pulled from all of your systems, now all of your partner's systems, and you get a trust report page to prove it to your customers.

[510] If you click the link in the show notes here or go to vanta .com slash acquired, you can get a free trial.

[511] And if you decide you love it, you will also get $1 ,000 off when you become a paying customer.

[512] Make sure you go to vanta .com slash acquired.

[513] So typically this would be where we would, you know, wrap up our history and facts.

[514] Companies just growing nice way.

[515] It's post -IPO, you know.

[516] Yeah, you know, we'd grade the IPO.

[517] We'd say like, man, you know, if only Apple had hold onto their shares and didn't need to liquid.

[518] It's kind of crazy.

[519] I'm not sure we would be doing this episode if SoftBank hadn't bought them because it is sort of like if you've bared with us this long in this episode, you probably agree.

[520] It's this sort of obscure technology company that has an absolutely inane business model.

[521] Before moving forward, David, it is worth putting a fine point on the business model where they don't manufacture chips.

[522] About half of their business or a little bit less, I think, comes from designing the chips and then licensing those out and then they have this large component of their business that is just licensing the instruction set architecture.

[523] They have now managed through the value they've created through the intellectual property but also the lock -in they've created by building an ecosystem around their instruction set.

[524] They just license their instruction set and it means they get a cut of all chips that are manufactured using their IP.

[525] And that sort of continues to blow my mind that somebody doesn't say, like, sorry, that's not your lunch.

[526] And this isn't like a trivial amount of money.

[527] I think it's something like, if you look at the iPhone, I think they make like 35 cents per unit.

[528] It's something like 1 to 2 % of the chip selling price, depending on which of their business models it was manufactured under.

[529] But like 35 cents of every, you know, A -series chip, why is it that they've managed to to hold on and actually get a cut of every single chip manufactured.

[530] I think it's probably two things.

[531] One, they continue to have truly excellent processor and chip engineering and design talent, right?

[532] So they truly are one of the world -class, you know, best processor design companies in the world at a time where advances in both performance and power consumption for processors are more important than ever.

[533] But I think also it's the same reason why the Wintel doopoly existed, right?

[534] Like even though Apple is more flexible about moving architectures than, you know, Microsoft was back in the day, Android and iOS run on risk architecture chipsets, are they really going to re -architect them for something else?

[535] There even is, there is no other thing to re -architect too.

[536] Yeah, that's true.

[537] You would have to come up with your own instruction set.

[538] build your own set of compilers to enable operating systems.

[539] So Apple does a lot of this anyway.

[540] Like Apple's one of the companies that actually could probably uniquely do it.

[541] Because they could write their own instruction set.

[542] I mean, this is like way bleeding into tech themes and random future forecasting.

[543] But like they could write a new compiler, build it into Xcode, make it so that every app that's compiled just works on their new chip set that's not armed base that they launch.

[544] Just like they did with Intel back in 2006.

[545] Right.

[546] And it wouldn't surprise me because they've taken over not only their own, I mean, they don't manufacture their chips, but they design their own ships.

[547] They of course license the ISA from Arm.

[548] But like it does just seem like it would be one more expanding of their vertical integration to say, actually, it's going to be based on our own architecture now too.

[549] So color me in for that prediction in the next three years.

[550] Interesting.

[551] And also Apple is uniquely positioned to be the only company to do it because Google's not going to do it with Android because Android is a diverse ecosystem.

[552] You need ARM to be the architecture to be the standard across the chip, you know, TI, Qualcomm, whoever, Broadcom, all the partners you need, plus Google, plus, you know, what have you.

[553] Google with Android is kind of the Microsoft.

[554] It's actually a little bit worse of a position than Microsoft because you can't guarantee that it's an Intel chip.

[555] It's probably a Qualcomm chip, but it could be a bunch of things.

[556] Well.

[557] All right.

[558] To pull back from tech themes, back to our story.

[559] Let's talk about the actual acquisition because that's the name of the podcast.

[560] this is like such a such a schizophrenic episode because until now I keep pulling us into the future this is so important this company and this technology and I hope we've done a good enough job communicating that but this is like a hardcore technology episode but now we're going to shift grab the wheel and shift back to like crazy you know James Bond style SoftBank and Masa.

[561] Okay so this brings us to 2016.

[562] All is well in the world.

[563] Arm is winning, you know, based in Cambridge, doing great engineering, working with partners, got their great business model.

[564] Did about $1 .6 billion in revenue the previous year.

[565] Yep.

[566] The stock price is having a huge run because the financial community is starting to wake up to this fact that they're like, oh, hey, if I want to ride the mobile wave and I want exposure to this, Arm is like a really good way to do it.

[567] Yep.

[568] We ship.

[569] 15 billion units in 2015 like it's a things are happening yep so thanks to fast company and Katrina broker and friend of the show David Lidsky and their great piece on Masa and soft bank that just came out we know what happens here so at summer of 2016 and Masa is hosting a dinner for tech industry luminaries at his amazing nine acre estate in Woodside California, which is like a super tony community, you know, right outside Palo Alto, right over 280.

[570] This is where Steve Jobs mansion was that he never built out.

[571] This is where Larry Ellison's whole compound is.

[572] And Masa has like an equally amazing compound there.

[573] He's hosting a dinner.

[574] One of the attendees at the dinner is Simon Seigars, who had joined Arm back in 1991, right after the spin -out.

[575] He was an engineer and he was the 16th employee.

[576] At this point, he's risen through the ranks over the years.

[577] He's become the CEO.

[578] And he's kind of like, you know, his job is, like, steward of this ecosystem.

[579] And like, obviously everyone uses arm, but like softbank is an important partner because they're one of the world's biggest telecom operators.

[580] And so, like, he wants soft bank to, you know, make sure that they understand how important arm is in the chipset of the phones that they're selling.

[581] Like, okay, great.

[582] I'll go to the.

[583] this dinner.

[584] I'll, you know, hang out with Masa.

[585] So they're sitting there at dinner and Masa starts like focusing on Simon and he starts asking him a bunch of questions about, you know, what arm does, what the business model is, you know, all the various devices that arm chips end up being used in.

[586] So Masa's like, so what exactly do you guys power?

[587] And he's like, well, we power everything, you know.

[588] We're not just in phones.

[589] Like we're in cars, we're in coffee makers, we're in refrigerators, wearables, we're starting to get into servers, like actually, because of this power consumption issue, like supercomputers, like power is a limiting factor on performance, so they're starting to use us.

[590] Masa's like, hmm, interesting.

[591] So then Masa asked him what, you know, now has famously become the question that he asks every Vision Fund investment, which is, what would you do if money were no constraint?

[592] and Simon's just sitting there at dinner and he's like um well I guess we'd probably keep doing everything but but faster and Masa's like hmm okay so Simon goes back to Cambridge back to the UK to Arm very different from the environment he was just in he's sitting there a couple days later back in Cambridge and he gets a call from Masa in Tokyo and in typical old Masa fashion.

[593] He's like, I need to see you right away.

[594] And not just you, but I also need to see Arm Chairman Stewart Chambers, Chairman of the board.

[595] And this is the best part of David's reporting here.

[596] This is great.

[597] Well, it's, we alluded to this in the carve out on the last show.

[598] So there's one problem, which is that Stewart is on vacation and he's on a yacht in the Turkish Mediterranean.

[599] But like, Maza doesn't like yachts in the Turkish med, like that.

[600] That's, that's his ballgame.

[601] Like, that doesn't phase him.

[602] He's like, okay, well, you know, hang on.

[603] I'm going to call him.

[604] He calls him and he's like, I need you to dock your boat in Marmé, Turkey, which is like a resort town on the Mediterranean coast in Turkey.

[605] Masa then sends a private jet to Cambridge, picks up Simon.

[606] He jumps on his own jet.

[607] They all converge in Marmé.

[608] I don't know if it's Marmé French pronunciation or Marmese in Turkey.

[609] Masa buys out an entire restaurant.

[610] with like a view overlooking the marina.

[611] So there's nobody else in there.

[612] They all converge the three of them in the restaurant.

[613] And they sit down and Mas says like, I want to buy you guys.

[614] And I'm going to offer you $32 billion, which was a almost a 50 % premium to where they were trading at that moment in time.

[615] And I want to do it now.

[616] You know, like when there's a public to public acquisition, it sometimes is like 20, 25%.

[617] We've seen up to 30 % and, you know, more typical situations.

[618] but like I think it's a 43 % premium.

[619] Yep.

[620] Typically with, you know, public to public acquisitions too, there's a lot of back and forth and there's investment bankers involved and like, you know, it's very rare that you get an Instagram style like done in it.

[621] They closed it in two weeks?

[622] Yep.

[623] They closed into, they announced, they got board approval and they announced within two weeks arm to remain an independent division within SoftBank.

[624] SoftBank's going to pay $32 billion.

[625] to buy the whole thing, the market loves it, it gets sharehold or approval, and it's done.

[626] This, as we alluded to at the top of the show, it was this model.

[627] I'm sure Masa was already starting to think about the Vision Fund, but this deal and this vision of computing everywhere and what it'll enable becomes really the prototype for what SoftBank is now doing with the Vision Fund, so much so that the next year in 2017, when they actually get the fund set up, soft bank corporate sells a 25 % stake in arm at cost to the vision fund.

[628] So the vision fund takes $8 billion right off the bat and buys 25 % of arm from soft bank corporate.

[629] So what do you think went on with that?

[630] Why?

[631] Well, I think Masa was always thinking about this as an investment.

[632] It was just that before the vision fund, the only way you had to invest was through soft bank balance sheet.

[633] Yeah.

[634] I mean, now there's a ticking time clock because the Vision Fund has a 12 -year fund lifetime.

[635] So it's not like he can hold it indefinitely now.

[636] Like at some point they have to either sell that back to SoftBank, which I'm sure Masa doesn't want to do as someone who controls both entities, probably wants to in the next, you know, five to seven years get a nice return on Arm and sell it to someone else.

[637] So we haven't seen the last of this yet.

[638] Well, I think most likely, and this is what people started talking about after the transfer, the 25 % transfer, yep, into the Vision Fund is re -IPO it.

[639] I think that's the most likely outcome.

[640] But what's interesting, and this also speaks to the Vision Fund strategy, so in 2016, when SoftBank acquired the company, there were only about 4 ,000 people working at the company, which on the one hand is a lot of people.

[641] On the other hand, like, that's a lot less than Uber.

[642] That's a lot less than, I think less than Airbnb at this point.

[643] Like for a company that old and that was like literally the foundation of, you know, all technology at that point.

[644] I mean, I think they have a 96 % market share of all smartphone and embedded devices, embedded devices.

[645] That's not a lot of people.

[646] In the, you know, two plus two and a half years since then, Arm has hired over 2 ,000 people.

[647] So they've grown by more than 50 % headcount since they couldn't have done that if they were a public company because they're now losing money.

[648] They're now net income negative.

[649] But Masa, again, like, you know, it's his question.

[650] What would you do if money were no object?

[651] Are they an income negative?

[652] So here's what I was looking at that.

[653] Their operating margin was 52%.

[654] Their operating margin at the time of acquisition, their operating margin now is 24%.

[655] Even though they've grown from $1 .6 billion in revenue to $1 .8 billion in revenue, their EBITDA dropped 40%, despite the revenue increase.

[656] So I'm not sure if they're losing money, but they, that's interesting.

[657] They're selling a lot more devices.

[658] They only make a little bit more money and their profits significantly dropped.

[659] Hmm.

[660] Interesting.

[661] And your margin dropped.

[662] I just pulled it up again.

[663] I said there's an article that they posted a loss of $200 million in, I believe, 2018, which could be for lots of reasons.

[664] Obviously, that's not operating margin.

[665] That's net income.

[666] The drum that they're beating and telling investors right now is, hold on, guys, we are way investing in.

[667] We are way investing in growth and it will pay off.

[668] And then we're going to do something about it when it does.

[669] And what's interesting, they still do quarterly presentations on company results.

[670] What they're talking about now is how they're investing heavily into the next computing waves.

[671] So Arm has started now making designs and devices chips that are dedicated for AI use cases, but also for like autonomous vehicle use cases.

[672] So you think about, you know, Intel bought Mobile Eye.

[673] for I think $17 billion, and MobileEye makes chipsets for cars, and in particular for computer vision for driver assistance and eventually fully autonomous use cases.

[674] Arm is getting heavily into that market as well, as well as all sorts of embedded computing.

[675] Yeah, it's interesting because the stated reason at the time of purchase, and at some point here we'll breeze through the other sections of the show, but at the time of purchase, the drum they were beating was it's IoT.

[676] that the Internet of Things is blowing up, and because, you know, there's going to be an arm device in much more than your phone, it's going to be in all these other things that are communicating with the Internet around you.

[677] That's why we believe that the growth is, it's like Masa believed that there was some growth that was not priced in that he believed was going to expand.

[678] The market for IoT devices was going to expand even greater than the public market investors who owned Arm believed, or at least that there would be some reason to believe that it would be generating more, more future cash flows.

[679] And it was IoT, IOT, IOT, and it's interesting to see, you know, two years later now on some of their investor relation stuff so that it really is more about the connected car and AI chips.

[680] So certainly their devices shipped has continued to be a really great story.

[681] They went from, I think, 2015.

[682] They were in, they were 15 billion.

[683] And now in, I think at the end of 20, when was this?

[684] 2017, they did over $21 billion.

[685] Yeah, that's what I'm looking at too.

[686] Their exponential growth continues to be really excellent, but it is interesting.

[687] It's like on what thesis, do you have to believe in a specific thesis about where there will be more CPUs, specifically low -power arm CPUs, or do you just say, look, I don't have to believe anything in particular.

[688] I'm just pretty sure that these things are going to continue to be more and more everywhere.

[689] Yeah.

[690] It's funny.

[691] I'm in the latter camp, but should we jump to acquisition category?

[692] Yeah.

[693] It's so funny.

[694] You know, we were setting this up all along for an IPO narrative, but boom, acquisition.

[695] Because we've done so many recently.

[696] I know.

[697] Business line.

[698] For sure.

[699] Not even.

[700] Yeah.

[701] I don't think SoftBank does anything else quite like Arm.

[702] So business line and it's fully independent.

[703] Okay.

[704] What would it happen otherwise?

[705] Like, you know, this is an interesting one.

[706] I feel like this is the first interesting one in a while.

[707] I have two interesting points to make on this one.

[708] So, do you know how Masa got the cash to buy Arm?

[709] Oh, interesting.

[710] I did not research.

[711] So it was interesting.

[712] The time when he bought it was actually at a currency fluctuation due to Brexit, where the pound was not doing well.

[713] So he got a lot of blowback that, you know, you're taking advantage of a distressed asset here, and he's pounding the table saying, no, I'm not taking advantage of a distressed asset.

[714] I really believe in this.

[715] And I was just waiting for the cash to come in.

[716] where the cash came in from was they sold $10 billion of Alibaba shares and $7 billion of supercell shares, plus they took a $9 billion loan.

[717] And SoftBank does this pretty often.

[718] I think they take out these big, big credit lines to do deals like this, which makes a lot of sense why you would switch to having a freaking huge fund if you're used to these multi -billion dollar loans to finance acquisitions.

[719] It's like, wait a minute, you pay me for using your money instead of me. paying you?

[720] I like this arrangement.

[721] It's kind of like the ESPN deal.

[722] It is.

[723] Business model innovation all over the place here it acquired.

[724] Yeah.

[725] So my sort of theme of this section is interesting to look at the conditions upon which the deal got done.

[726] So that's condition one, which is interesting to look at.

[727] The second condition is they were a nice deep pocketed Switzerland that this company could actually sell to that could finance aggressive future growth.

[728] And if you think about other people they could have sold to, it would have been value destructive because if you sold to Apple, then, you know, Qualcomm's freaking out that they're not going to have access to their core technology anymore.

[729] And same thing in, you know, the other direction.

[730] And so the number of possible purchasers who have $30 billion or who can raise $30 billion that are willing to plow a bunch of future investment into your business that aren't strategically or I guess structurally sort of corruptive to where you want to operate in the market in a value destructive way, pretty limited.

[731] I would say it's fortunate and value creating for the world that it landed where it did.

[732] Well, yeah, I mean, think about like, we referenced at the top of the show that in many ways this became the model for the Vision Fund.

[733] And, you know, look, say what you will about the Vision Fund, and there are certainly, you know, like anything, there's a balance sheet, right, of like it's their positives and negatives to the ecosystem of it.

[734] You know, this is a major positive, right?

[735] Like, Arm is a public company, right?

[736] It had been a profitably operating public company for at that point close to 20 years.

[737] It was the most, arguably the most important foundational computing company out there.

[738] But like if they were to say to Wall Street, hey, you know, we're going to massively invest in growth right now we're going to turn net income negative you should expect losses for the next few years while we lay the groundwork for the next waves of computing what would happen right like their stock would crash like they they were like going to be prevented from doing that and i think this is the core of you know in some ways it's like a hilariously funny and masas question of like what would you do if you know money were no object but this is like this is what he means right like and uh i think in a lot of ways like soft bank is a great home for this company for the time being.

[739] David and I have gotten a lot of feedback on, and listeners, we appreciate all this feedback, that we do a lot of talking about the value that accrues to the acquirer, but we do very little talking about the value created in the world by a transaction happening.

[740] This, I think, is value creating to the world if you believe that those 4 ,000, you know, really brilliant physicists and PhDs and chip designers and computer scientists and technologists, if you believe that funding them, they'll continue to produce IP that will enable us to have continued innovation and do things that we previously didn't think possible because it would produce too much heat or, you know, consume too much energy or anything like that, then yeah, I think this is incredibly value creative both for the arm ecosystem and for sort of anybody who uses any of their products, which is all of us.

[741] Mm -hmm.

[742] Mm -hmm.

[743] Totally.

[744] it's it's very much a you get the investors you ask for there's a very reasonable chance that the public market was not a good place to uh you know to go stand on that hill and say we're going to be we're going to be bad for a while but it'll be great eventually it's like you need someone else to sign up for that right right it's very hard to ask you know once you have a certain set of investors it's very hard to ask them something different from what you have been asking them without a without a change of control like this yep Should we move into tech themes officially?

[745] Yeah, let's do it.

[746] Finally.

[747] I've got two, which I know I've been beating this drum for a while.

[748] One is just like this alignment that I talked about earlier of like technology model and business model.

[749] Like if you can really innovate on both of those, then like that is when really, really, really powerful things happen.

[750] You know, maybe in the theme of this episode and Apple and its involvement.

[751] here.

[752] You know, it is the marriage of technology and the liberal arts when you can do that.

[753] So that's one.

[754] The other one that I feel like I've been talking about more on the LP show.

[755] I can't remember how much I've talked about this on the main show that I think this really illustrates very well is when it comes to startups and investing, it's so easy.

[756] People make the mistake of focusing on what your TAM is today versus the much more important question of what your TAM is tomorrow.

[757] And this is such a great case study because like when Arm the spinoff was getting started and they were developing the risk processor, you could have looked at it and been like, this is so stupid.

[758] Like the TAM, you know, the market today is for IBM PCs and it's the Microsoft, you know, Intel, Wintel duopoly is getting started.

[759] like this company is dead in the water like this is so stupid right but it was the tam tomorrow of the coming mobile wave that that was the opportunity and that was what mattered now of course that means that getting the timing right is super important but if you can time correctly when the tam tomorrow is going to start to realize and you can get out in front of that and be building the technology you know a couple years in advance so that it's ready like that's that That's magical.

[760] I am looking at a graph here with about 10 years where it looks like there's no growth in revenue.

[761] It's timing, timing, timing.

[762] But it wasn't that long until from the time of the spin -out, they got that TI contract for the Nokia 61 -10 within three years of the spin -out.

[763] So the devices weren't shipping yet, but they knew it was coming.

[764] Yep.

[765] That's a great point.

[766] I have no tech themes.

[767] I have said all of mine.

[768] Love it.

[769] Love it.

[770] Love it.

[771] For grading, what do you think about?

[772] We've talked about for grading more recent events.

[773] And I think this still qualifies as a recent event, the SoftBank acquisition.

[774] Let's paint the A -plus scenario and the C -R -C -minus scenario over the next five years.

[775] All right.

[776] So I'll start with the A -plus.

[777] if you believe in the thesis that David and I chatted about on the LP show that that ambient computing is the next computing wave so to sort of review the there was PC that there was a trillion dollars in that ecosystem of value created then there was the internet and then mobile and sort of what is that next thing and what is that next wave of computing upon which platforms will be built that you know that enable entirely new use cases, new companies, and, you know, a new way that people interact with technology, and ambient computing is that thing, they're probably all going to have arm chips in them, be it the AirPods that Dave and I are both wearing now, or the watches that I think we're both wearing now, or as our friends and the team that work there, say, the lady in a tube that will respond to my voice when I get home, all the way down to sort of the litany of things that used to, you know, just be dumb embedded devices like the little oven timer or the little oven clock, there's a very reasonable chance that we have hundreds of arm chips that we're interacting with over the course of the day.

[778] That makes us look like a pretty smart buy.

[779] Assuming that that was not already priced in.

[780] So when we just, like, let's review that real quick.

[781] The $32 billion was bought for 18x earnings and 29x EBITDA.

[782] So already a, you know, fairly expensive stock, but yeah, still feels like a good buy if you believe that we're going to transition from sort of somewhere between one and a dozen that we regularly interact with to hundreds.

[783] Yep.

[784] And it's funny.

[785] I'm just pulling up here because of course Arm still reports their results.

[786] It's awesome that they do that.

[787] I know.

[788] It's so awesome that they do that.

[789] Okay.

[790] So totally agree.

[791] That's the A plus.

[792] case the c minus case uh i think i fall in the a camp but just to paint the bare picture here so currently for the most current results we have total revenue for arm over the last four quarters so the trailing year of revenue was 1 .8 billion so they bought the company for soft bank bought the company for 32 billion two and a half years ago for a company that is still doing less than two billion in revenue um that's a pretty big gap to fill right like if you're anticipating like you know significant returns on uh on your acquisition from you know already two and a half years in the past you need some of these new markets to start hitting and start hitting big real soon from uh from an irr perspective and you know the royalty portion of the company's revenue, which is the alignment with partners.

[793] So that's the one that's dependent on number of units shipped.

[794] The one that's dependent on number of units shipped.

[795] And that's really what, you know, I think the investment thesis is about that, like number of units of shipped is going to grow exponentially.

[796] Like, it is the largest revenue stream in the company, but it's actually not growing that fast in terms of how much revenue they're getting.

[797] And I think that must be because, they're just getting such a small amount from every device shipped.

[798] And as the devices proliferate and cheaper and more basic devices, you know, like sensors and the like become the vast majority of number of units, I bet they're getting fewer actual dollars from those.

[799] Yeah, it's interesting.

[800] And this is an area, one other bare case is something that I'm not technical enough to evaluate.

[801] But what Arm was to mobile and Intel was to the desktop is there, going to be something else to the proliferation of devices and make arm sort of look like the the old grandpa technology like as there's some significantly lower power thing that you know literally just needs the vibrations of the em spectrum in the air to power it or something really crazy now that's why i think like this is in the full length of time like i don't see anything else on the horizon that can actually really replace arm and the and the risk architecture but again if you look at their financial results, like the fastest growing portion is software and services, right?

[802] This sounds like the Apple narrative, right?

[803] Of like, you know, all as well, even though our, you know, core business is flat to declining.

[804] Like, we're making it up in services.

[805] Like, that's not as defensible.

[806] So, you know, I think that's this, this bear case is the company, while a great company and a foundational technology company actually can't command that much.

[807] of the value creation going forward from computing and doesn't end up being as good an investment as they expect it to be.

[808] From the academic side of things, it depends how much arm is the point of integration.

[809] If they're just a component that ships in phones and the real value is created, and the phone is the point of integration of hardware, software services, and components, then most of that value accrues to the phone manufacturer, which is why Apple's worth or half of trillion dollars.

[810] but to the extent that Arm actually has ecosystem lock -in and commands a lot of value for doing that, and in some way they're the point of integration, bringing the software developers who are able to write software that gets compiled to run on Arm processors with the actual device manufacturers, which they aren't.

[811] That explains why they, you know, command so much less of the value.

[812] There you have it.

[813] We shall see.

[814] We shall see.

[815] Man, this is a fun one.

[816] Thank you, listeners, for bearing with us.

[817] I hope we did justice and communicated, you know, a little bit about the technology, a little bit about the business model, hopefully a lot about the story.

[818] But also, you know, why this is, no matter what corner of the technology ecosystem you live in, this is an important one to understand.

[819] Yeah.

[820] And I don't think I really got that until digging in.

[821] I mean, it was always kind of mysterious to me of like, I just didn't understand why it got bought for so much when they don't make anything.

[822] They weren't making the chips.

[823] And it's, you know, it's been nice to have an hour and a half to understand that a little bit better.

[824] Our sponsor for this episode is a brand new one for us.

[825] Statsig.

[826] So many of you reached out to them after hearing their CEO, Vijay, on ACQ2, that we are partnering with them as a sponsor of acquired.

[827] Yeah.

[828] For those of you who haven't listened, Vijay's story is amazing.

[829] Before founding Statsig, VJ spent 10 years at Facebook where he led the development of their mobile app ad product, which, as you all know, went on to become a huge part of their business.

[830] He also had a front row seat to all of the incredible product engineering tools that let Facebook continuously experiment and roll out product features to billions of users around the world.

[831] Yep.

[832] So now StatsSig is the modern version of that promise, and available to all companies building great products.

[833] Statscig is a feature management and experimentation platform that helps product teams ship faster, automate A -B testing, and see the impact every feature is having on the core business metrics.

[834] The tool gives visualizations backed by a powerful stats engine, unlocking real -time product observability.

[835] So what does that actually mean?

[836] It lets you tie a new feature that you just shipped to a core metric.

[837] in your business and then instantly know if it made a difference or not in how your customers use your product.

[838] It's super cool.

[839] Statsig lets you make actual data -driven decisions about product changes, test them with different user groups around the world, and get statistically accurate reporting on the impact.

[840] Customers include Notion, Brex, OpenAI, FlipCart, Figma, Microsoft, and Cruise Automation.

[841] There are like so many more that we could name.

[842] I mean, I'm looking at the list, Plex and Versel, friends of the show at Rec Room, Vanta.

[843] They literally have hundreds of customers now.

[844] Also, Statsig is a great platform for rolling out and testing AI product features.

[845] So for anyone who's used Notion's awesome generative AI features and watched how fast that product has evolved, all of that was managed with Statsig.

[846] Yep.

[847] If you're experimenting with new AI features for your product and you want to know if it's really making a difference for your KPI's Statsig is awesome for that.

[848] They can now ingest data from data warehouses.

[849] So it works with your company's data wherever it's stored so you can quickly get started no matter how your feature flagging is set up today.

[850] You don't even have to migrate from any current solution you might have.

[851] We're pumped to be working with them.

[852] You can click the link in the show notes or go on over to stat sig .com to get started.

[853] And when you do, just tell them that you heard about them from Ben and David here on Acquired.

[854] Well, folks, if you aren't subscribed and you want to hear more, from your favorite podcast client.

[855] If you like what we're talking about about kind of the future of computing and ambient computing, you should consider becoming a limited partner.

[856] So Kimberlite .fm slash acquired or click the link in the show notes.

[857] And we'll have some pretty fun guest, guests coming up on the LP show pretty soon that we're excited about.

[858] So we will see you all soon.

[859] Indeed.

[860] See you next time.