Once, I thought running a hardware company would be all about making hardware. Turns out, it’s actually about learning new things.
When we started back in 2006, I was an embedded software engineer: I knew how to crank out high-performance C and assembler for mobile phones, games consoles, and microcontrollers, but absolutely nothing about how to design chips or PCBs, still less how to build millions of units of a physical product and get them to customers all over the world.
Still, it’s amazing what you can pick up if you hang out with bright people.
This article first appeared in The MagPi 75 and was written by Eben Upton
Once upon a time
Around the time I was building the first Raspberry Pi prototype (on Veroboard!) in 2006, I joined Broadcom as a software engineer in the video and graphics team, and got my first taste of hardware design. Although I was supposed to stick to writing firmware, every software engineer had access to the Verilog source code for the chip that we were working on. Nobody in their right mind would let me write a module from scratch, but if I could find a bug, and get my fix signed off by the real hardware engineer in the next cubicle (future Raspberry Pi COO James Adams), my code might make it into the chip.
A couple of years later, we started work on the first real Raspberry Pi design. There’s a vast difference between soldering down some PDIP chips on a piece of Veroboard, with signals running at a few tens of megahertz, and escaping high-speed signals like HDMI from a fine-pitch BGA on a six-layer PCB. But I got to watch over my co-founder Pete Lomas’s shoulder while he laid out what would become the classic Raspberry Pi Model B. And six months later, I got to watch Mike Buffham (future Raspberry Pi Director of Product Management) figure out how to build Pete’s design in volume at Sony.
I’ll never be able to write Verilog like James, lay out a PCB like Pete, or juggle a supply chain like Mike. But hanging out with them taught me more about hardware, and about business, than any course or textbook could.
Going to print
Which brings me to publishing. At the start of 2015, we had the opportunity to take up the reins of The MagPi from Ash, Will, Ian, and Aaron, who had been running it since 2012. Liz was quite insistent that we should do this, and equally insistent that we should recruit Russell Barnes as editor. After five digital-only issues, the magazine made a triumphant return to print in August 2015, and has never looked back. Russell has surrounded himself with an amazing team, including Lucy and Rob here at The MagPi, and Ben and Andrew at HackSpace, and together they’ve done amazing things.
I spend a wholly inappropriate amount of time thinking about our publishing business. In part, that’s because it’s further from my comfort zone than anything else we do: there’s just so much for me to learn here. But it’s also because even I can see the scale of the opportunity ahead of us. I think you’re going to love what the team has planned over the next couple of years, but for now let’s raise a glass to the first 75 issues of The MagPi, and to the next 75.
