I'm Materials Science student interested in using an RPi cluster for calculations of materials properties.

I'm very new to HPC but what I know about the RPI is that its processor is based on the ARM architecture which was initially intended for control and now we see this architecture on phone processors, but I don't know the limits I can push this type of cluster to, with heavy calculations. I also know that they will function as a means to devide the calculations in threads with the propiate parallel programing so I will not ask if they work as a single virtual machine, I understand it doesn't.

My question is if the RPi cluster performance ratio, while stacking, does keep increasing for heavy math calculations or any physics calculations; or if there is a point where the performance curve gets steady.

If anyone have used a RPi's clusters for Physics, Math or Chemistry calculations, how does it performs?

Thanks in advance.

1 Answer 1


If you're interested in learning the basics of how the RPi functions and how to do such tasks, the RPi is a great tool.

That being said, an RPi cluster has no chance in hell at defeating even a moderately recent i5 processor. Your laptop will probably serve you better than an RPi cluster for heavy computations.

  • Seconding this. RPI is great for a project to run a sensor network and to offload computation directly to a sensor. It can also be supremely useful to connect an older research tool to a network and allow remote data collection. However, it is not a serious computation platform, it has an impressive amount of power, but it is $35 after all . The most reasonable use I see for a true cluster is redundancy for long term installations.
    – crasic
    Dec 23, 2017 at 3:23
  • The main reason for asking was the price. I could stack 50 of them and enhance some of my work, and it would be way cheaper than a mixed CPU/GPU cluster. But if you are saying it is a no go, well I guess I'll have to invest in a more serious one. Thanks for the reply guys.
    – dbarcene
    Dec 23, 2017 at 16:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.