How Raspberry pi Started, It's future plans & much more | Talking Tech with Eben Upton

I'm feeling fantastic. I'm feeling somebody who got off an airplane about 12 hours ago. I'm feeling surprisingly good. I'm feeling awake, which is nice. I've had some coffee. I don't want to think about what time it is in the UK right now. But no, it's great.

It's good to be back, right? It's good for me personally to be back, and it's great to have the sense that the maker, the whole maker scene, is back now, very firmly back after the pandemic, after the supply chain crisis. You can buy Raspberry Pis again. You know, all of it, it's like it was before, but better.

Oh, wow. Um, so, a tiny little thought. The original thought was a tiny little thought, which is like, 'Hang on a second. Young people used to be excited about computers, and huge numbers of young people used to get excited about computers at home, and then they used to then apply to study computer science at university, and then suddenly we woke up one day, and there were no young people anymore. Where had they all gone?' Right? And so, really Raspberry Pi is just to test a hypothesis. And the hypothesis is there used to be programmable computers, BBC Micro, Commodore 64, Sinclair Spectrum, and that was how kids got excited. Those machines went away, and then the kids went away. If we bring the machines back, will the kids come back? And that's all we're doing. That's our Raspberry Pi is just a big out-of-control hypothesis test. And we were just trying to increase the number of people applying to the university in Cambridge. Computer science used to be, in 2008 when we started, computer science was the easiest subject to get into at Cambridge. We had barely two to three applicants per place. That's an incredible ratio, right? And that's the problem we were trying to solve. Interesting thing, last year we had 1,800 applicants for about 100 places, right? It was the hardest subject to get into at Cambridge. So, and we're not the only people who've contributed to that, but there is a sense that that mission, that kind of mission accomplished. As we accomplished the one mission, we started to see more and more and more missions.

Yeah, we're an educational, and our owner, our top co, is our own Raspberry Pi Limited. So, we run the engineering organization. Our owner is the Raspberry Pi Foundation. So, a UK charity, a UK not-for-profit. It does an enormous amount of work actually now in India. And has 18 employees, I think, in India. You know, dedicated to getting young people excited about knowledgeable about computing.

I think it's more or less the same. I mean, the fascinating thing, as you say, 70, 80% of our business is now industrial. You know, and the idea, yeah, yeah, yeah. So, the educational business is way larger than we imagined because we thought, 'Oh, a thousand Raspberry Pis, 10,000 Raspberry Pis.' So, we're selling a couple of million Raspberry Pis a year into education and hobbyist communities, but we're selling five or six million Raspberry Pis a year into industry. The fascinating thing for me is there's never been a moment where I have had to make a different decision about a main role, a core Raspberry Pi products and SBC product because of industry or education. The things we do for one community always help the other.

This is the single biggest change in Raspberry Pi in the commercial part of Raspberry Pi, is the move from being purely an electronic products company to a blend of an electronics product company and a semiconductor company. We launched RP2040 in January of 2021, amazing time to launch a microcontroller when you couldn't buy anybody else's microcontroller, so you know that's a bit of luck there. Yeah, very very low prices. This is why, you just wait. And the other really important thing, I've also been inspired by SpaceX here actually, because SpaceX you can go on their website, you can still go on their website I think, and they'll tell you what a rocket launch costs. You can't go on Boeing's website and Boeing won't tell you what a rocket launch costs, you have to negotiate with them. I'm sure you know, SpaceX say our rocket launch costs this much money, you write us a check, we'll launch you a rocket, right? That was just inspirational, this idea that you know the semiconductor pricing in the semiconductor industry. We have people working for us at Raspberry Pi who used to work for mainline, you know, broad line semiconductor companies in sales and marketing, and they used to have tools to help them price their product for a particular customer that was designed to milk as much money as they could out of every customer. My idea is like, well, I'm just going to make this chip, I'm going to tell you what it costs. RP2040, it costs 70 cents if you buy 3400 of 13-inch reel, it costs 80 cents if you buy 500 units on a 7-inch reel, and it costs $1 if you buy one of them, right? That's it. That's the story. Now if you go to our, you know, we sell it through distribution. If you go and try and get one of our distributors to sell you a million units, I'm sure they'll do you a deal, right? But, you know, but that's none of my business. That's none of my business. So, low pricing, transparent pricing, we're not done with low pricing yet, either. So, you know, so what was wonderful for me as a maker was that there were maker companies that were going to go bankrupt, right? In 2021, 2022, when they couldn't buy silicon, we kept maker companies alive, right? When they were abandoned by ST, when they were abandoned by NXP, when they were abandoned by Microchip. These companies who owe some of their success to the maker movement abandoned the maker movement. We tried to stay true, we tried to the best of our ability through the crisis, we tried to stay true to the maker movement, particularly to the maker pro world. These people, you know, we talk about, you know, with the big Raspberry Pi products, we had to, we said we prioritized industry during the shortage. When people think industry, they think multinational companies. Yeah, sure, we sell Raspberry Pis to multinational companies. But most of our business by, you know, they're startups. It's hundreds, it's a hundred units, it's a thousand units. It's a married couple who mortgaged their house to start a business. Right, what are we going to do? We're going to cut them off? You know, we like, you know, so we tried really hard and RP2040, because it was in unlimited supply at a point where the big Raspberry Pi products were in shortage, that was kind of like that was the backbone of our ability to make a contribution.

Yes. Yes. Yes.

We don't talk about unannounced products. Well, I did give you some insider information, which is the pricing. So, the pricing trajectory is important to us. I think we're going to continue to try to optimize the price of the existing products. There's a, as to whether we're going to do more semiconductor products, there's a science fiction novel called Cryptonomicon by an author called Neal Stephenson. And there is a point in the book where somebody challenges the main character. She says to him, 'I think they're doing something like laying fiber optic cable in the Philippines, right?' And she says to him, 'Are you going to lay some more fiber optic cable?' And he says, 'People in business seldom plan to do something once because it screws up the spreadsheets.' I don't want to have screwed up spreadsheets. So, you've seen in the last month you've seen another silicon product from us, RP1, RP1, right? Okay, so clearly, if people's model was that I hired a bunch of semiconductor engineers, made RP2040, and then fired them all, clearly that's wrong. Okay, so I obviously have a semiconductor team. They're obviously doing work. I think you can expect to see interesting new semiconductor products from us. And you know, in the reasonably near term now.

I think it's really important to emphasize that there's a huge difference between building a microcontroller or even building something like RP1. RP1 is a very complicated chip, but there's a huge difference between building these devices that are on relatively old process nodes. They're all on TSMC's 40 nanometer process node. That was the process node we started on at Raspberry Pi back in 2012, right? There's an enormous difference between the central processor that we get from Broadcom for Raspberry Pi 5 is a 16 nanometer product. There's a huge difference between building relatively simple chips on relatively old processes and building advanced chips on new processes. So, you know, I wouldn't hold your breath for that. We have some fantastic partners in this space. But you know, where we can make a contribution, microcontrollers. You know, you're going to see us invest in that as much as possible. And you know, you may well see us, there may well be an RP1.5, you know, there may well be future that architecture, that chiplet architecture where we've decoupled fast digital logic in BCM2712 from analog and interfacing silicon in RP1. Yeah, you'll see us probably evolve both sides of that chiplet split.

Yes. Yeah, yeah, yes, yes, yes. They're the same, they're the same thing actually, because they are both entrepreneurial, entrepreneurism and makerism, right? They are both about not waiting for permission to do things, right? So yes, I'm a person, I've always been a person. I have a 3-year-old son and he doesn't wait for permission to do things. I look at him and I'm like, 'Oh god, here we go again, right?' I remember what this must have been like for my parents. Yeah, I've never waited for permission to do anything in my life and that's made me a maker and it's made me an entrepreneur. And you know, the fact that the maker community and the entrepreneur community have co-evolved and have come together and have blurred together in the middle there in that kind of maker pro world that maybe we and Seed, I guess, are the two most extreme examples. I just did a great session with Eric Pan. We and Seed are probably the most extreme examples of that coming together of the maker and the entrepreneur community. It is wonderful and it's given me an opportunity to spend my adult life not asking for permission.

Yeah, thank you.