Description

I own Raspberry Pi.

I have long admired Ubiquiti (my last VIC writeup) which is a fantastic albeit unusual company whose stock is up about 1,500% since its IPO in 2010 (~24% IRR). Here’s an overview of the how and why behind the company’s success.

Ubiquiti was founded by Robert Pera who was an engineer working at Apple in their networking group which made routers. Robert is an engineer and he despises typical corporate trappings like middle management, meetings, sales and marketing, etc – he believes the right products will eventually just win. Robert has created an entrepreneurial culture unlike just about any other company of any real size whereby employees are expected to innovate and take full ownership over the products they work on. Ubiquiti makes inexpensive products that are easily accessible to customers and cheap enough so that people see little downside in trying them. The products tend to be disintegrated and free of extraneous features and are designed to maximize performance for one particular application. Ubiquiti also makes their own software to manage the hardware, and they give it away without hefty licensing fees like competitors do. It’s a flat organization where titles don’t mean anything and executing is all that matters. The company is run by engineers, for engineers. It is unapologetically not for everyone. This has been the right combination of organizational structure and product philosophy to create the innovation which has caused Ubiquiti to consistently develop disruptive and successful products in multiple product categories. When Ubiquiti IPO’d in 2010, the company had $137 million in revenue, almost all of it in their internet service provider business. At the time, the company had just gotten into the networking business, which no one believed could be successful as they were going up against companies like Aruba and Cisco. Today, that segment does well over $1.5 billion in sales. These companies evolve and Ubiquiti has been profitable every year, though what you bought in 2010 is not what you see in 2024. I see a lot of parallels.

March 2005

All the best hackers I know are gradually switching to Macs. My friend Robert said his whole research group at MIT recently bought themselves Powerbooks. These guys are not the graphic designers and grandmas who were buying Macs at Apple's low point in the mid 1990s. They're about as hardcore OS hackers as you can get. The reason, of course, is OS X. Powerbooks are beautifully designed and run FreeBSD. What more do you need to know? I got a Powerbook at the end of last year. When my IBM Thinkpad's hard disk died soon after, it became my only laptop. And when my friend Trevor showed up at my house recently, he was carrying a Powerbook identical to mine. For most of us, it's not a switch to Apple, but a return. Hard as this was to believe in the mid 90s, the Mac was in its time the canonical hacker's computer.

Raspberry Pi is ubiquitous in the tech world yet virtually unknown outside of it. They make computers and related computing hardware. They also make their own operating system. So, it’s a computer company, except their niche is that they make tiny computers called single-board computers (“SBC’s”). Raspberry Pi makes other computing devices as well, for example they make microcontrollers (“MCU’s”) which are sorta like SBC’s except they’re a bit dumbed down. Whereas an SBC is a full computer that can handle a multitude of tasks, an MCU is more specialized for one type of application. Their SBC’s and MCU’s come in different flavors for different needs (generally there’s a price vs performance tradeoff) and Raspberry Pi has a broad lineup of different products. Raspberry Pi also makes peripheral devices such as cameras, touch screens, touch sensors, power adapters, keyboards, cases, antennae, AI kits, etc. Again, it’s a computer company, and they’ve sold about 70 million units to date. They keep growing the number and types of products.

In the fall of 1983, the professor in one of my college CS classes got up and announced, like a prophet, that there would soon be a computer with half a MIPS of processing power that would fit under an airline seat and cost so little that we could save enough to buy one from a summer job. The whole room gasped. And when the Mac appeared, it was even better than we'd hoped. It was small and powerful and cheap, as promised. But it was also something we'd never considered a computer could be: fabulously well designed. I had to have one. And I wasn't alone. In the mid to late 1980s, all the hackers I knew were either writing software for the Mac, or wanted to. Every futon sofa in Cambridge seemed to have the same fat white book lying open on it. If you turned it over, it said "Inside Macintosh." Then came Linux and FreeBSD, and hackers, who follow the most powerful OS wherever it leads, found themselves switching to Intel boxes. If you cared about design, you could buy a Thinkpad, which was at least not actively repellent, if you could get the Intel and Microsoft stickers off the front.

Okay, so what’s special about this company? Raspberry Pi designs products to be disintegrated, which allows customers to only pay for what they need, without being forced into buying more expensive products that incorporate seldom used parts and features. Costs of products are kept very low by using things like commodity parts and older process nodes for chips. They are not trying to compete with typical computing companies, instead they are playing almost entirely at the low end of the market…this is not for data centers. Raspberry Pi puts very little money or effort into sales and marketing. Their entire S&M budget is basically to cover half a dozen salespeople for larger industrial accounts and technical writing and handbooks for their products. Raspberry Pi has a massive community of users who are active in posting their projects and helping with build ideas and troubleshooting. It’s actually Reddit’s second-largest subreddit among tech companies after Apple with 3.2 million members.

Raspberry Pi has moved beyond just assembling commodity third-party parts and now has their own in-house chip design capabilities. So far, the company has successfully designed and implemented three chips, the RP2040 and its successor the RP2350 (both are used for MCU’s) and the RP1 which is an I/O controller for the Pi 5, their most recently launched SBC. They also played a role in developing the Pi 5’s processor which is made by Broadcom, using their own IP for GPU and video capabilities. Raspberry Pi sells the MCU chips and collects royalties on the Broadcom chip for external revenue, though they don’t sell the RP1 to external customers. ASP’s for the MCU chips are about $1 and they’ve sold over 10 million to date. They have a particular focus on blending performance and power consumption. I think it’s clear that Raspberry Pi will continue to improve their chip design capabilities and move up the curve to make more advanced products.

The company also makes its own OS called simply Raspberry Pi OS, which is used on about 2/3^rd of all Pi’s. The OS is optimized for Pi’s and customers seem to like it. They put a lot more effort into hardware, but I suspect software (it’s open source) will be an increasing area of focus for them.

With OS X, the hackers are back. When I walked into the Apple store in Cambridge, it was like coming home. Much was changed, but there was still that Apple coolness in the air, that feeling that the show was being run by someone who really cared, instead of random corporate deal-makers. So what, the business world may say. Who cares if hackers like Apple again? How big is the hacker market, after all? Quite small, but important out of proportion to its size. When it comes to computers, what hackers are doing now, everyone will be doing in ten years. Almost all technology, from Unix to bitmapped displays to the Web, became popular first within CS departments and research labs, and gradually spread to the rest of the world.

Speaking of the Apple store in Cambridge MA, in 2001 I was living in a studio apartment across the street from the Cleveland Circle MBTA stop outside Boston. I was and still am a music fanatic…it was and somewhat still is an obsession in my life. I had a Diamond Rio MP3 player at the time, which was as good as it got for MP3’s. I didn’t know much about Macs except for the fact that I typed all my papers (and did my fantasy basketball drafts) in my college’s computer lab stocked full of those translucent rainbow-colored Macs. Anyway, I came across Steve Jobs’ presentation of the first-generation iPod and was floored. Like, literally mouth agape. So, that weekend I took the green line to the CambridgeSide Galleria – the same one Paul Graham was at – which was the only Apple store for hundreds of miles. The sales guy tells me that the iPod only works on Mac OS, meaning I would have to buy a ~$2,000 Mac if I wanted a $500 iPod. Screw it. I had practically nothing to my name, I made $24,000 a year and had a Capital One credit card with – and I’m not kidding – a $100 credit limit since it was my first credit card. It was a glorious product…by the way, I still have it with the original lightning cable, and it still works albeit with only two hours of battery life and a song catalogue that’s frozen in time from 2007. I rode the green line to work every day, and I immediately began getting stopped up to ten times per day by people asking what my iPod was and asking all sorts of questions. It was constant. One day I was at the gym (the Boston Sports Club at the corner of Mass Ave and Newbury, which is apparently now a Planet Fitness) and local boxing legend Pete “Hurricane” McNeely was there doing some sort of boxing clinic. I was in the locker room and Hurricane, naked as the moment he was born, approached me with all these questions about the iPod. He was soooo into it. So, I had this ten minute conversation with a fully naked palooka who got absolutely destroyed by Mike Tyson and probably couldn’t half-complete a children’s crossword puzzle but man was he super interested in my iPod. Anyway, the light bulb went off and without opening a single 10K or anything else, I took almost every cent to my name, including some money that my grandmother had left me, and put it all in Apple stock in my Datek account. (It was the only brokerage that didn’t have an account opening minimum) Because I’m stupid, I sold it all in 2007, but as Charlie Munger said, “That first $100,000 is a bitch, but you gotta do it.” Charlie was right…thanks, Apple. Thanks for indulging me, I enjoy this story and hopefully it has application to this writeup.

Okay, back to Raspberry Pi which was founded in 2010 by a group of computer scientists at Cambridge University who had an altruistic vision to spread their passion for computing, engineering, programming, and hacking. They thought they could accomplish this by making very small, very cheap computers that were underpowered by today’s standards but perfectly capable of handling education and creative projects. Their idea took off and their first SBC sold out quickly (10,000 units with a $4 ASP) and I don’t think it’s unfair to say that there’s been customer obsession ever since and today they sell something like 7-8 million units annually. Originally, the company was a charitable organization though it’s obviously now converted to a for-profit company, and today the charitable foundation (which still has the same education-focused mission) owns half of the company. I know this sounds corny but I’m glad about the foundation as the company has a bit of a soul to it.

What’s cool about the evolution of this company is that it started out being used (outside of the education market) almost entirely by engineers, developers, and hackers at home. As these people tinkered with their projects and fell in love with Raspberry Pi, many of them realized they could use this stuff in their day jobs at whatever companies they worked for. This process has been slow but it’s taken hold and today about 75% of Raspberry Pi’s ~$300 million in revenue is now attributed to commercial and industrial customers while the rest is for the hobbyist and education markets. What are these being used for? Tons of stuff, a lot of IoT-type things where you’d never know there was Raspberry Pi powering it. Pi’s are great for smaller production numbers are prototyping, however I believe that companies will increasingly ditch custom hardware even for larger production runs and stick with Pi’s as they simply work too well for the cost, or they’ll use custom hardware that’s got Raspberry Pi chips in it. The market for these things should continue to expand.

I remember telling my father back in 1986 that there was a new kind of computer called a Sun that was a serious Unix machine, but so small and cheap that you could have one of your own to sit in front of, instead of sitting in front of a VT100 connected to a single central Vax. Maybe, I suggested, he should buy some stock in this company. I think he really wishes he'd listened. In 1994 my friend Koling wanted to talk to his girlfriend in Taiwan, and to save long-distance bills he wrote some software that would convert sound to data packets that could be sent over the Internet. We weren't sure at the time whether this was a proper use of the Internet, which was still then a quasi-government entity. What he was doing is now called VoIP, and it is a huge and rapidly growing business. If you want to know what ordinary people will be doing with computers in ten years, just walk around the CS department at a good university. Whatever they're doing, you'll be doing.

Hardware is a notoriously difficult business as products get copied and margins get eroded. One could look at Raspberry Pi and assume they’re merely assembling commodity off-the-shelf parts; shouldn’t this business get competed away? In fact, a quick search turns up the fact that there is no shortage of competitors to perhaps be concerned about. For starters, large computer OEM’s such as Asus and Dell have made attempts at competing with their own SBC’s, although none have generated measurable sales and most have given up – not a bad sign. There are a couple companies that specialize in SBC’s and have some real sales, but they’re Chinese so they’ll never get any real traction outside Asia. Among these, the only two companies that have been able to create businesses from SBC’s are Rockchip Electronics and Espressif Systems. Their products are good but no Western company will touch it for commercial applications when Raspberry Pi is sitting right there. (Hobbyists still use these companies though) Finally, many Raspberry Pi clones have copied the “Raspberry” name and call themselves other types of fruits like “Orange Pi” in hopes of competing. It reminds me of around 2004 when every hardware OEM tried competing with iPods, often with a name that tried to rip off “iPod.” Sony, Dell, Creative, Sandisk, and Microsoft all released perfectly fine products that died quick deaths, unable to compete against the incumbent. None of these Raspberry Pi clones have made a dent. Even if you’re unsure of the exact reasons for Raspberry Pi’s ability to keep thriving, there’s clearly something that’s caused this company to be successful despite allegedly making cheap commodity hardware. This is one of those times where you have to ask yourself the question: if this is a commodity business, why doesn’t anything about its financial results resemble one?

In the matter of "platforms" this tendency is even more pronounced, because novel software originates with great hackers, and they tend to write it first for whatever computer they personally use. And software sells hardware. Many if not most of the initial sales of the Apple II came from people who bought one to run VisiCalc. And why did Bricklin and Frankston write VisiCalc for the Apple II? Because they personally liked it. They could have chosen any machine to make into a star. If you want to attract hackers to write software that will sell your hardware, you have to make it something that they themselves use. It's not enough to make it "open." It has to be open and good.

Eben Upton was one of Raspberry Pi’s founders and he is the CEO and basically the face of the company. Eben is a great product guy who seems to understand running a business and the importance of making money. He looks like Jason Statham if he got super into Dungeons & Dragons. Raspberry Pi has been profitable at least each of the last three years, with operating margins in the low teens. This is not your typical profitless tech company hoping to ramp things up one day. There are many public interviews with Eben where you can get a good sense for what drives him, what his values are, what kind of manager he is, and what his expectations are for the company. My takeaway is that as far as running a company goes, Eben is very serious about growth and profits. He very much understands that in order for the foundation to continue to pursue its goals, the company has to thrive. Since the early days he’s been extremely cost conscious, and I believe this is today embedded in the company’s culture.

On valuation, Raspberry Pi trades at 21x my estimate of forward P/E. Is this cheap? On the surface, I guess not really. It’s not nosebleed but it’s also not optically cheap…kind of that no-man’s land where without any thinking it’s hard to get fired up but cheap enough to get interested. I will make a couple comments about the valuation.

First, while lumpy, the company has shown impressive growth that continues today. The recently reported 1H24 results (the company reports in USD but the stock trades on the LSE) showed revenues up 61% and EBITDA up 55% largely on the strength of the Pi 5 launch. What is the addressable market for Raspberry Pi today? Hell if I know. Large…I’m betting larger than people realize. But again, I expect it to be lumpy.

Second, I believe Raspberry Pi falls in the bucket of “expensive for a reason” IPO. In fact, this has been a hunting ground of mine for a while after doing some research that showed that many stocks with +20% IRR’s while going public in the last few decades IPO’d with P/E multiples north of 20x. Obviously, there have been plenty of underperformers with the same high multiples. At some point every stock involves a leap of faith.

Third, innovation has always been incompatible with value investing because it’s difficult to assign a tangible value to something so inherently vague. Yet, it’s arguably the most valuable part of many successful companies. We’ve seen this over and over again. Why is Amazon a multi-trillion dollar company? Because they stopped “selling books” and evolved into a logistics behemoth, then just sort of invented a business out of thin air that happened to be an early mover in the next generation of computing architecture. I could go on but my point is that it’s always been a mistake to overlook highly innovative cultures that have records of success inventing new stuff and usually by way of a benevolent dictator founder are committed to keeping the structures that top talent – who aren’t wired like most of us – are drawn to.

What do I think it’s worth? It’s difficult to say – is “significantly more over time” too laconic? I can only think of a few comps that can at least point us in the right direction in the short-run. The two Chinese companies Rockchip Electronics (603893 CH) and Espressif Systems (688018 CH) are each basically Raspberry Pi clones. Rockchip has similar revenues as Raspberry Pi ($400 million vs $300 million for Raspberry Pi) though with higher margins. It’s mainland Chinese though, which means I couldn’t explain its valuation (7.5x sales, 36x EBITDA, 50x P/E) which is about 2.5x Raspberry Pi. Espressif has about 10% less in revenues but again, trades at over twice the valuation as Raspberry Pi. If anything I would think this would be flipped but what do I know. Ubiquiti is another comp, and its shares have historically traded between 20-30x P/E, in the “expensive for a reason” category. The last comp is Samsara (IOT) but at 20x sales and 180x P/E I don’t take it seriously as a valuation signpost. Comps suggest higher at the present.

And open and good is what Macs are again, finally. The intervening years have created a situation that is, as far as I know, without precedent: Apple is popular at the low end and the high end, but not in the middle. My seventy year old mother has a Mac laptop. My friends with PhDs in computer science have Mac laptops. And yet Apple's overall market share is still small. Though unprecedented, I predict this situation is also temporary. So Dad, there's this company called Apple. They make a new kind of computer that's as well designed as a Bang & Olufsen stereo system, and underneath is the best Unix machine you can buy. Yes, the price to earnings ratio is kind of high, but I think a lot of people are going to want these. – Paul Graham

Am I being grandiose by interspersing this prescient Paul Graham article from 2005? Yeah. To be clear, Raspberry Pi is not on a trajectory to become a $3 trillion company, but it’s still a useful exercise to understand how these things can take off. There’s something here. I think a $1 billion valuation for Raspberry Pi looks quite attractive.

Catalyst

Can we get rid of the catalyst section?