I’ve assembled a list of useful single board computers (SBCs) for use in homelabs. They are essentially low power computers useful for lightweight tasks and experimentation. I’ve used them (and still do) for certain tasks, especially when I’ve wanted to experiment with various things on the ARM64 platforum.
- FriendlyARM NanoPi M4 (6 ARM64 cores, 2/4GB LPDDR3/LPDDR4 RAM) – a really useful board for storage applications. It features a PCI Express interface, allowing for multiple SATA ports using the SATA hat or an NVMe drive using the NVMe hat. Both configurations are supported under Linux. I’d personally use Armbian as the Linux distribution rather than the official releases. I’ve written a tutorial on how to use it as a NAS with ZFS earlier.
- ODROID HC1/HC2 (8 ARM cores, 2GB LPDDR3 RAM) – useful as a storage node due to the SATA port (over USB 3.0 instead of PCI Express compared to the NanoPi M4 above, though), or for multithreaded compute uses due to the 8 core Samsung Exynos ARM CPU.
- NVIDIA Jetson Nano Developer Kit (4 ARM64 cores, 4GB LPDDR4 RAM, 128 GPU cores) – single board computer useful for development in artificial intelligence and machine learning. Due to the lack of GPU power it’s mostly useful for inference however. Another option for similar use cases could be the Google Coral single board computer or the Google Coral USB dongle. I’ve written a tutorial for this board on how to make it run in headless mode.
- Raspberry Pi 4 (4 ARM64 cores, 1/2/4 GB LPDDR4 RAM) – the most well known single board computer. Useful for most low load server applications. The greatest thing about the Raspberry Pis compared to other single board computers is the community around them. The Raspberry Pi 3, Raspberry Pi Zero and Raspberry Pi Zero W are also great choice depending on the use case.
I’m working on more tutorials and articles relating to these and other single board computers. Hopefully they’ll be ready soon. 😊