Pi in the sky? Having sold 1million devices worldwide, the Raspberry Pi Foundation looks to the future

5 min read

Since its launch in February 2012, Raspberry Pi – the credit card sized computer – has been something of a phenomenon. Demand for it has crashed websites and those who have since got their hands on one are exploring all manner of potential applications.

The question which many are asking is where the device goes next? And that's a question which the Raspberry Pi Foundation isn't going to answer – at least in the short term. Pete Lomas is a co founder of the Foundation and its hardware guru. "We're not saying anything about the future because we don't want to raise people's expectations," he claimed. One thing is certain, however: the Foundation will be looking to build on the device's success. But the success has come in areas where it wasn't expected. "It's blossomed in some unexpected areas," Lomas admitted, "but not in some areas we targeted, including education." And he thinks education will continue to be a hard nut to crack. "It's a difficult sector to address," he said. But this is being addressed. The high level of Pi sales has brought working capital to the Foundation and that has allowed the Foundation to appoint a director of educational activity. "It's a full time job to get that running." And the educational goals have been helped by Google, which has donated 15,000 Pis for UK schools. One problem is convincing the education sector that Pi is more than just a board. "We have to demonstrate that the value of using a Pi to learn how to debug is not just a single skill. I think debug is a multifaceted analytical process." He contends that engaging students with software debug helps to develop analytical skills. One of the difficulties, he contends, is in getting students to think for themselves – and that was where the idea for the Pi originally came from. "I've interviewed graduates who haven't reached that stage," he recalled. "They often can't bound a problem. Take multiplying 100 x 100; the answer isn't going to be less than 100, neither is it going to be more than 1million." Lomas is all in favour of getting people to think for themselves. "What excites me is the 'what if?' factor. What's an approximation? The ability to divide problems into chunks is a great skill to have and it acts as a 'pull through'. In some respects, it's like off roading; you're not sure what you will have to deal with, but it's more exciting." Track Raspberry Pi back to its origins and you will come across Raspberry Pi Foundation trustee Eben Upton. At the time, he was director of studies in computer science at St John's College Cambridge and responsible for undergraduate admissions. "He realised there was a problem," Lomas noted. "Students were coming to university with no debug or code writing skills. The college was having to 'bootstrap' students and the first term was effectively lost." Upton came up with the idea of giving students a computer in June and telling them to do something useful with it. When they started their course, they would be expected to 'show and tell'. "Upton built a prototype in 2006," Lomas noted, "but that was as far as it went." Fast forward a couple of years and Lomas became involved with the initiative, by accident it appears. "Norcott (Lomas' company) had built some electronics for Imperial College and I came to London to see it working," he said. "I met with Professor Alan Mycroft from Cambridge University and when we were walking through Hyde Park, we got on to the topic of students. Upton's idea came up and we both decided it was brilliant and that we should build some. I went to Cambridge, met with the team, saw there was critical mass. I thought 'we can do this' and had a go." But things didn't go quite as planned. "The original device was designed around the Broadcom 2727 running the Python programming language – that's where Pi comes from, by the way. We couldn't hit the price target because there were too many barriers and the device itself didn't quite do enough." A further problem was device packaging; the Broadcom chip came in a 0.4mm pitch bga, so stuffing the board was a challenge and would have increased costs further. "It didn't quite go on the back burner," Lomas said, "but we had to wait for technology to catch up." Upton – now with Broadcom – then suggested the Broadcom 2835. It came with an ARM core, a gpu and was packaged in a 0.65mm bga. It also allowed memory to be mounted within the package. "Because we could put 256Mbyte on top, that pushed us into thinking about Linux," Lomas said. Interest in the Pi was building. Dave Braben had demonstrated a device on YouTube and that got 600,000 hits. "We knew there was interest," Lomas said, "so we had a shakedown to get it done." At first, the Foundation didn't expect to sell too many, but YouTube changed that. "The original idea was to make a couple of thousand and to subsidise them, but we realised we needed a sustainable business model." Meanwhile, the Pi website was launched, offering Model A for $25 and Model B for $35. "The prototype board was costing more than $100," Lomas said, "so we stripped it back to the necessities and hit the target price." But that meant having Model B made in China. "That wasn't our desire," Lomas reflected. There was also the small problem of sourcing the components needed to build the boards. Raspberry Pi almost didn't happen. "We needed a manufacturing strategy," Lomas recalled. "Eben put his business degree hat on and we talked with Farnell and RS about licensing the design and building it. They took it on and the rest is history. Without their sign up, we would still be hand soldering boards and wouldn't be where we are today." Then, when the Pi was launched, websites crashed. "We were criticised for our inability to deliver, but we couldn't put 1m units on the shelf before hitting the go button." Since the launch, there has been some work on the software and a few issues with the Model B have been fixed. The lower cost Model A is available and manufacturing is now being performed in the UK by Sony. "Sony has been brilliant," Lomas pointed out, "and its manufacturing error rate is in the basement." But it's the future which is of interest, particularly to the rapidly growing body of Pi enthusiasts. "Pi has now got a life of its own," Lomas believes. "The ecosystem is driving itself and the supporters are doing a great job of helping each other. They're telling us if things are going wrong and so on. It's a community which we never expected, but which we value immensely and recognise the hard work being done. And magazines and books are appearing which support it." Maintaining Pi's progress is important. "What we don't want is for Pi to be a one trick pony," Lomas said. "The educational goal is too important. So our focus continues to be on the Pi's educational value. How can we improve that? Would the addition of more computing power increase the educational value? Would the introduction of more accessories make a difference?" One accessory that is about to be launched is a camera module. "It's a key move," Lomas contended. "The chip has a powerful gpu and we want that to be used for educational purposes – there's no point in having a gpu if you can't use it." Lomas' view is that vision will be a key learning experience. "People need to learn how facial recognition works, for example," he said. "The module will bring a new level of opportunity." So will there be a Pi2 and, if so, when might it appear? "We don't have a mantra that says we have to produce another model every six months so people have to buy it," Lomas noted. "We have always said that we'll work hard not to disenfranchise the people who bought a Pi on day one. Any improvements we make will be as backward compatible as possible, with a mind to support early adopters." Nevertheless, Lomas admitted the Foundation is thinking about Pi2 – what it might be and how it might fit with Pi1. "Moore's Law says there will be a better processor and there will be more memory and that will allow us to maintain the price." And Lomas is concerned about the price. "We need better worldwide distribution," he said, "because a Pi Model B doesn't cost $35 in Brazil; once the taxes and so on are added, it's more than $100. We have to find a way to get it available at lower cost." One thing you can't accuse Lomas of being is pessimistic. But there is one thing about Pi which saddens him. "We've sold a lot of Pis," he concluded, "but I'm always despondent when I hear of people who bought one because of the hype, but haven't done anything with it since."