---
title: '🥼 BOINC'
url: 'https://www.niklas-stephan.de/en/blog/boinc'
markdown: 'https://www.niklas-stephan.de/en/blog/boinc.md'
lang: en
date: '2022-11-01'
description: '🥼 BOINC If you can be enthusiastic about scientific projects on one hand and have a bit of computing power to spare on the other, you should take a closer look at the open-source project and tool BOINC. In this article I report on what BOINC actually is and how you can install it easily. Voluntee…'
taxonomy:
  category:
    - science
  tag:
    - boinc
    - distributed-computing
    - open-source
---

# BOINC

   1st Nov 2022   [boinc](https://www.niklas-stephan.de/en/blog/tag:boinc#body-wrapper) [distributed-computing](https://www.niklas-stephan.de/en/blog/tag:distributed-computing#body-wrapper) [open-source](https://www.niklas-stephan.de/en/blog/tag:open-source#body-wrapper)  

If you can be enthusiastic about scientific projects on one hand and have a bit of computing power to spare on the other, you should take a closer look at the open-source project and tool BOINC. In this article I report on what BOINC actually is and how you can install it easily.

![](https://www.niklas-stephan.de/user/pages/blog/boinc/boinc-certificate-2022-04-01-2048x1371.webp)

## Volunteer computing?

> Volunteer computing describes a technique of application programming in which individual computer users make computing capacities such as CPU time and storage space available to a specific project on a voluntary basis, in order to compute a shared result using distributed computing.
> 
> — Wikipedia

That means anyone who can spare a bit of computing power can participate in scientific projects to help them solve certain problems. To do that, you merely have to install a specific program that runs on a computer operated as often / continuously as possible. So anyone who, for example, runs a Raspberry Pi at home or – like me – a server anyway, can do scientific work by providing parts of their free processor capacity.

## Note

Before running it on a laptop or a classic PC, consider that increased CPU usage reduces the lifespan – perhaps not significantly – and increases power consumption. Running it on cloud server offerings such as Microsoft Azure or Amazon AWS is also inadvisable because of their billing models based on CPU performance and data transfer.

— Otherwise it can get expensive!

## What is BOINC?

The creators of the BOINC tool set themselves the task of making around 30 scientific volunteer-computing projects available and configurable through one interface. The tool also provides settings – such as the resources to be provided – and reports on work already contributed. An exact list with descriptions of the projects can be found here: <https://boinc.berkeley.edu/projects.php>. It also becomes clear there that not every project exists for every computer platform, and others are not suitable for every kind of platform. Some of them require, for example, a fast dedicated graphics card, since the computations involved are optimized for that processor type (GPU).

![](https://www.niklas-stephan.de/images/4/0/d/c/3/40dc3332f9cc4cafb6b7e437b1495a3328cb2a95-02-boinc-screenshot.webp)

## Okay, I like it. How do I start?

Installation can either be done manually and directly via the installer available on the BOINC homepage, or – as I prefer and describe below – as a Docker container with Docker Compose.

The manual installation works on every kind of computer and can be configured, for example, so that BOINC runs as a screensaver and thus only claims the PC's resources when it is not being used. You can download the installation from <https://boinc.berkeley.edu/download.php>. For operation on a server in a container, here is an example `compose.yml` file:

```yaml
services:
  boinc:
    image: lscr.io/linuxserver/boinc
    hostname: "boinc"
    container_name: boinc
    environment:
      - PUID=1000
      - PGID=1000
      - TZ=Europe/Berlin
      - PASSWORD=xxx
    volumes:
      - ./config:/config
    restart: unless-stopped
    logging:
      options:
        max-size: "10m"
        max-file: "3"
    networks:
      - dmz
    deploy:
      resources:
        limits:
          cpus: 1.00
          memory: 2048M
networks:
  dmz:
    external: true
```

Specifying the resources provided via `deploy – resources – limits` is optional. A restriction of the resources provided to the program can, as already mentioned, also be configured directly in the program. I play it safe here and restrict access from the outset. A CPU value of 1 corresponds to one CPU thread. With, for example, 8 threads present, the container defined here can access 12.5% of the available CPU performance. Exactly the "background hum" I want. Specifying the port is only required if we do not use a reverse proxy; then our BOINC instance would be reachable on the host via the corresponding port. The password should definitely be set and is then required to log in to the virtual desktop via the browser. The user, by the way, is hard-set to `abc`.

## Usage and conclusion

Either by starting the application or by accessing the web interface in container mode, we can now sign up for the projects we are interested in within the program and start the processes. After that nothing exciting happens anymore, and further intervention on our part is not required. We can check our contribution to the respective project at any time via the reports integrated into the program or on the respective project page. There we can also have a certificate generated, if we want.

It cannot be simpler – have fun (letting it) compute!

---

## Navigation

- Parent: [Blog](https://www.niklas-stephan.de/en/blog.md)
- Previous: [IR-Remote Control 4.0](https://www.niklas-stephan.de/en/blog/ir-remote-control-40.md)
- Next: [ncms](https://www.niklas-stephan.de/en/blog/ncms.md)
