So why is the Raspberry Pi foundation, the organization behind this charming device, making the computers in the first place?
We spoke with Eben Upton, executive director of the Raspberry Pi foundation to find out why. Here’s what we learned:
- It’s primarily intended for the education market. The whole idea was conceived as a way to get kids to learn how to manipulate and program computers earlier on.
- The Raspberry Pi foundation wants to open-source the technology so “a company in China can produce a million computers” for developing countries and schools. The foundation expects third parties to start developing Raspberry Pi devices midway through 2012.
- The multimedia performance of the Raspberry Pi is “substantially better” than the Tegra 3, a chip used in many modern smartphones, Upton said. The only smartphone that comes close to the Raspberry Pi’s performance is the Galaxy S 2, he said.
- They don’t intend to make money off it. While you could easily turn something like this into a fully operational business, the Raspberry Pi foundation will remain a not-for-profit, Upton said.
- Around 10,000 units should be available once or twice a month. There’s an upper limit of about 100,000 that the Raspberry Pi foundation can produce in a year, though.
And here’s the full interview:
BUSINESS INSIDER: Why did you guys want to build such a cheap computer?
Eben Upton: We came up with the idea because we’d been interviewing potential undergraduates to come to Cambridge university about 5 years ago. Both the number of people applying and the stuff you could have relied on them already done was getting worse. The numbers were going down and could hardly rely on the people you did get to know anything about computers.
We looked around for reasons why this happened. The thing that came to me, the people of my generation had small computers when they were kids. They had TRSATs, they had these machines and they were programmable. You turn them on and the first thing you could do was print “hello world.” These are going away and have been replaced by game consoles or PCs, which are programmable.
I started looking for a way you could provide a machine cheaply enough that you could give you children, settling on this $25, $35 price point. Over the last 5 years, we’ve been looking at ways for making a machine like this. I joined Broadcom and it turns out Broadcom made chips that ware really cheap. You could build a pretty respectable computer at the $25 point and the foundation is really an organization that brings out the possibilities of this.
Now you’ve got a chip that can meet the price point, the foundation is a way to do that.
BI: So it’s a shot at getting kids to learn how to program?
EU: Yeah, they’re so cheap you can give them to all the children or they can buy them like they buy textbooks. That’s the idea, children are enormously illiterate now, but what they know how to do is use computers. They see them as bits as functional magic and have no idea how they work. That’s fine for Facebook and browsing, but if you want a career out of this stuff or create something that’s high value, you have to understand how the thing works
This is almost nationalist. We were concerned about Cambridge’s problem and the university’s problem of getting enough qualified students. Then we were concerned about Britain’s problem, not producing enough engineering graduates. It was a quite parochial initial view we had. AS soon as news got out that we were going to do this, most of the interest we saw was in the undeveloped world. Russia and Brazil, a lot of people very interested in this.
The project has broadened out from this educational thing to adult hobbyists. A lot of the biggest cheerleaders are guys my age who want to build robots and media centers. Also people in the developed world where you can get performance out of places with televisions but not computers. It turns your TV into a workable productivity computer.
BI: Why show off the video game performance of it, then?
EU: I guess what we tried to do, we showed you running a web browser, a piece of productivity software. We wanted to emphasize everything the chip can do. It’s a maddeningly powerful process, it will run a desktop. It won’t set the world on fire with its desktop performance, but it has a lot of multimedia performance. It can do 1080p HD video playback. We wanted to put out a series of videos capturing it doing these things that surprise people at that price point.
BI: Do you guys ever plan to make money? Or turn this into a business?
EU: There is no corporate organization. The Raspberry Pi foundation has six trustees, I’m the executive director of the foundation. The foundation owns all the intellectual property embodied in the device and is the business entity procuring the manufacturing and handling distribution. It’s a limited company under English company law. It’s possible to take a company like that and register it as a charity. The company is registered as a not-for-profit.
The money you get is recycled back into the business. The bearers of the trustees have given loads of money to the foundation. That provides the working capital required to pay for chip infantry. The primary limit on our scale is the working capital to hold our infantry and buffer it as it runs through the company, we have pretty insignificant fixed overheads.
We’ve raised capital in 10,000 unit batches to build the devices. That’s the money we need, that will provide us, but there’s an upper limit to how many devices you can build in a year at that rate. With best use of working capital you can build 100,000 devices each year, to scale we’ll have to raise additional capital. We’re intending to release the designs for the device at due cost. We can’t make any money out of this, we have no incentive to keep the design of the device secret.
We do hope third parties will be able to manufacture clones. We can expand the concept without having to expand the capital base.
BI: Does that mean you guys are planning on releasing a second version?
EU: We’re comfortable with our multimedia performance, we do realize our ARM performance is kind of retro. 700MHz is enough, but it could be more, but we don’t currently have any plans for a successor.
Obviously we’re careful not to speculate. We’ll see how this one goes. The history is full of computer companies that have imagined the next product and talked about it and then people have fallen out of sway. There are no concrete plans. Look at Apple, it’s a company that I really admire in a lot of ways and they are extremely good at controlling information and their image. We’re going to do that.
BI: So when are you taking reservations for the device?
EU: We actually haven’t taken pre-orders. We built an initial prototype batch, alpha boards, we’ve had those for several months. We’ve built a very short run of the final device, a test run to make sure the design is sound. It does appear to be sound, we found one small design defect, it’s a five minute fix and we’ve fixed that now. We’re in the process of committing a manufacturing run. We’ve bought parts for 10,000 devices and we are in the process of committing a manufacturing run for that.
When those devices come back in a few weeks time, we have a web store that we’ll turn on. We’ll turn the storefront on with a few thousand devices for sale. I suspect it’s gonna take an hour to sell through it at that point, I’m going to hold a few hundred in reserve. We have developers we’ve committed priority devices to.
We’ve been unusual in not taking pre-orders, a lot of people try to fund the capital requirements of the project by taking pre-orders. I think we could do that, but it’s always felt very risky to me. It creates the risk of, if something goes wrong, we’re going to end up having defrauded a lot of people a lot of money. Even now, with a workable device, we’re still very careful about not taking people’s money until we have a physical device in our hands.
BI: How often do you expect to have Raspberry Pi computers available?
EU: I suspect we can do a batch slightly more than once a month. Looking at the supply chain, it will require some careful planning to do that within our capital requirement. I think that’s pretty achievable. The aim very early on is to get these designs into the hands of the parties. We would like nothing more than some company in China to make a million of these. It would be perfect, we would achieve our goal, which is ubiquitous presence of cheap computers without having all the requirements of doing capital raising to scale.
Fingers crossed. I hope that’ll be in the first half of 2013, clones will be a factor.
