The idea of an open source laptop makes me feel the same excitement I felt using Linux for the first time over a decade ago.
Earlier this month Andrew "bunnie" Huang, hacker of the Xbox (read the book) and creater of the Chumby announced a crowd funding campaign for his Novena open-source laptop. I'd heard about the project before and liked the design. In fact I have several almost identical system level designs in my engineering notebook waiting for me to have the time and money to make one myself. Since it is unlikely I'd ever get around to building anything that costly or complex on my own I'm really excited to hear that I can buy one of someone else's design.
I'm really excited by the idea of an open source computer that is more than just an under-powered toy. A computer that I can use for day-to-day coding and layout, running a normal full distribution of Linux. Some people have questioned what the point of having it open source is when no one can reasonably go about fabricating a one off for themselves. The reason is documentation. Just like software if you have a closed source black box that doesn't do what you expect based on the documentation you're stuck. If you can get at the design source you can figure out what is going on and maybe even make it do something that the designer hadn't originally anticipated. Having the schematics and design files makes it possible to find a place to hook into the system management bus, or where you can get power for your custom wireless adapter. The idea of an open source laptop makes me feel the same excitement I felt using Linux for the first time over a decade ago.
There's a large Spartan6 FPGA on the motherboard, tightly coupled to the ARM CPU. This is a feature I'd designed into my own idea for an open laptop and makes a whole world of exciting and interesting possibilies available to the hobbyist. How about a custom hardware graphics accelerator? A software defined radio, custom crypto engine, physics co-processor, massively parallel modular synthesiser... The options seem more or less endless, and the hard bit with a lot of those projects is the high speed signal integrity hookup to a decent general purpose processor for network and storage, which is already done for you in this board. I've worked on the Ettus Research USRP E100 software defined radio in the past which uses a similar arrangement for streaming radio samples into the internal ARM processor. Combining this very fast data transfer with the processing capabilities of the FPGA and the ARM processors NEON vector floating point unit made it possible to do in real time what it was taking a quad-core Intel i7 was barely keeping up with. Like all application specific accelerators it only works if you have a lot of repetitive processing to do, but there are so many compute intensive jobs that I'd like to tinker with I really like this feature.
This is probably the part that has attracted the most criticism from around the web. Only the top price $5000 laptop is a traditional clamshell based design. The others are all in one monitors really. I tend to agree that the "laptop spec" mdoel isn't useable as a laptop, which is a little disappointing but it's out of my price range anyway. The case certainly has some innovative and hacker friendly ideas in it, but it isn't a laptop. The argument that you can open the screen up and put a keyboard in-front and use it like a laptop only works if you've got space. As I write this blog post I'm on a bus so no room for a separate keyboard here.
I have a feeling that some of the connectors and pinouts used in this design are going to rapidly become standards amongst the open source hardware world I've already seen a software defined radio board adapted to fit the Novena high speed FPGA interface. I'm thinking of using the battery charger interface connector in my next project so that my project can use the open source battery management board from the project, or I can use my own battery charger design with the open source laptop. This kind of standardisation for things like big lithium batteries and high speed connectors is a good thing and will take a high profile, high value project like this to get started.
I'm getting one, I really look forward to using it and hacking it to add features or solve problems I couldn't with a closed laptop. The stretch goals all involve investment in more open hardware so the more funding the faster this area will develop.