PicoCluster – A ready-to-use Raspberry Cluster

picocluster Raspberry

PicoCluster offers the possibility to use our Raspberry boards (or other SBCs) to create a low-cost distributed parallel system.

You are the lucky owner of a set of Raspberry Pi 4, Jetson Nano, Google Corel, Odroid C4-U4, RockPro64, Radxa Rock PI 4B or Up! Squared X86 type boards. They are still working boards, maybe from the previous generation… how to reuse them?

Here is the idea: let’s create a nice cluster of boards for distributed computing!

Moreware has already had the opportunity to explain what PicoCluster represents. They offer complete “turnkey” cluster systems based on SBCs, or assembly kits for makers.

With PicoCluster it seems to be back to the early days of home computing. These small Clusters (from 3 to 20 boards) are not super powerful, but that’s okay. For a large number of use cases at home and in the company, they are more than powerful enough. The best part is that they are inexpensive and a 10 node cluster only requires 30 watts of power instead of 3000 watts. Think about it!

Not sure how you can use PicoCluster? Here are some ideas. These are 101 projects that have successfully used the PicoCluster system.

Examples:

  • Run Docker containers
  • Deploy Kubernetes clusters
  • Set up a load-balanced web server cluster
  • Create a database cluster
  • Create an AI/ML training cluster
  • Host a personal cloud storage
  • Set up a CI/CD pipeline
  • Run a multi-node Hadoop cluster
  • Create a private VPN server
  • Host a Git server

These ideas can be adapted to your specific needs and explored further based on your use case.

PicoCluster Raspberry

The versatility of PicoCluster makes it suitable for a wide range of applications, both personal and professional.

All the assembly instructions can be found on the blog. This page acts as an index of the blog to make it easier to find specific instructions.

For the most part, all the assemblies are very similar in construction. The Raspberry PI 5 models have several versions, but the assembly is generally similar. The Raspberry PI4 4GB and 8GB, ODroid C4 and Rock PI clusters are almost identical, while the ProckPro64 clusters are slightly different due to the different board sizes and layouts.

PicoCluster also offers excellent online support, which is ideal if you run into any difficulties.

 

Subscribe to our Telegram groups

Useful Links

Follow us to not miss the next news!

Definire ciò che si è non risulta mai semplice o intuitivo, in specie quando nella vita si cerca costantemente di migliorarsi, di crescere tanto professionalmente quanto emotivamente. Lavoro per contribuire al mutamento dei settori cardine della computer science e per offrire sintesi ragionate e consulenza ad aziende e pubblicazioni ICT, ma anche perche’ ciò che riesco a portare a termine mi dà soddisfazione, piacere. Così come mi piace suonare (sax, tastiere, chitarra), cantare, scrivere (ho pubblicato 350 articoli scientfici e 3 libri sinora, ma non ho concluso ciò che ho da dire), leggere, Adoro la matematica, la logica, la filosofia, la scienza e la tecnologia, ed inseguo quel concetto di homo novus rinascimentale, cercando di completare quelle sezioni della mia vita che ancora appaiono poco ricche.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.