It’s easy to be dismissive of the humble Raspberry Pi. In many ways it’s painfully limited by slow I/O, meager compute and a reliance on an microSD card to boot. But despite these shortcoming, and perhaps because of its bargain basement price, the board has found a hoard of devotees. Keith Townsend looks at how the Pi could find a home in the data center. He makes a good point, despite being low power, the compute on it is like my favorite pizza, “cheap and deep”. Keith thinks this could have major applications as an object store, citing OpenIO as one company who has a specific walkthrough for it.
I think the more interesting application is in IoT. Keith describes a Raspberry Pi being used as a monitor for other devices, and something that could be used as a general scripting node. As a general purpose SoC, its value decreases the more you try to shoehorn it into what it wasn’t designed for. But in these settings, where the programability of the Pi is the most important feature, it can make a lot of sense.
The Raspberry Pi makes sense in the data center as long as you view it as a platform rather than a product. Its functions in and of themselves aren’t very impressive. But when you can put that platform virtually anywhere, with a fully capable OS, it changes how you think about it. The Pi’s versatility far outweighs its cost. That’s the reason it’s taken off with consumers, and that’s why Keith sees it having value in the data center
Keith Townsend comments:
Despite its CPU and I/O limitations, the Raspberry Pi potentially represents the future of computing with interesting options in the enterprise data center.
- MacOS Has Been Around Longer Than Classic Mac OS - May 24, 2018
- Jon Hildebrand – IT Origins - May 24, 2018
- Q(uestion)LC Flash | Gestalt IT Rundown: May 23, 2018 - May 24, 2018
- Pure Accelerate Live Blog – Day 2 - May 24, 2018
- Pure Accelerate Live Blog - May 23, 2018
- BONUS: Patric Palm – IT Origins - May 22, 2018
- HPE Buying Plexxi - May 21, 2018
- Random Numbers on a Raspberry Pi - May 17, 2018
- FPGAs are Coming for AI Workloads - May 17, 2018
- Patric Palm – IT Origins - May 17, 2018