Testing and comparing small-board computers

The Rasp Pi exemplifies a classic success story. Since the Model B came on the market in February 2012, there has been no stopping the little computer. As of October 2015, more than seven million models carrying the Raspberry Pi label had been sold. The Rasp Pi has attained these heights of success even though it was neither the first nor the most powerful small-board computer (SBC). Its selling points are an attractive price tag of $35 and the GPIO (general-purpose input/output) header, behind which the Raspberry Pi Foundation educational mission lies: Teaching young people about hardware and software programming through tinkering and experimentation.

Nobody knows exactly how many projects have thus far been implemented with various Rasp Pi models, but the numbers probably run into the millions. These projects range from the smallest to clusters of circuits to the Astro Pi project [1], which put two Rasp Pis on the ISS space station a few months ago so that school children could carry out experiments they themselves had designed. The Raspberry Pi and its competitors are also catching on in the fields of security, automation, and robotics. Additionally, they are finding application as multimedia centers and as small components of network-attached storage. However, these types of uses are bringing the Rasp Pi, even in its third generation, close to the limits of its performance capabilities.

Success stories such as that enjoyed by the Rasp Pi always attract copycats, especially from Asian companies that have been busy releasing competing products on the heels of new Rasp Pi developments. These offshore companies hope to gain a foothold in the booming small-computer market. However, the hardware itself is not the only feature attracting consumers to the Rasp Pi. Its $35 price tag is hard to beat.

Equally important is what the Rasp Pi has achieved in just a few years: a well-functioning infrastructure with well-maintained software and accessories and a large and active community. Efforts at copying these types of developments have fallen short in the Asian commercial space in part because their manufacturers do not necessarily place the same kind of emphasis on open source thinking. For example, Allwinner, which manufactures semiconductors in China continues to have a negative effect on the GPL [2].

In spite of its strong position in the marketplace, the Rasp Pi is hanging on to a concept that puts it at a disadvantage with competitors. The Raspberry Pi Foundation wants the board design to remain downward compatible, and it is difficult to compensate for the limitations of this approach. For example, the RPi3 introduced at the end of February 2016 only has a 10/100Mbps Ethernet connection, even though most of its competitors already offer Gigabit Ethernet. The necessary remedy here would be to change the Rasp Pi layout from the ground up since the Ethernet connection in the present design depends on a slow USB 2.0 controller.