BOINC Central provides computing power to scientists at universities and research labs, and to independent researchers. Their jobs run on thousands of volunteered home computers.

What we support

• The computation must have a scientific or non-profit goal.
• The software must be open-source, as defined here.
• After the research is completed, its conclusions must be made public, ideally in a journal or conference paper.
• Input data must not contain personal or sensitive information.
• Input and output data must be publicly available, unencrypted, on demand.

Technical guidelines

• The computing workload should consist of large numbers of independent jobs. Volunteer computing provides high throughput, not low turnaround time.
• Jobs should run on typical home computers: they should use no more than 16GB RAM and a few GB of disk space.
• Job runtimes should be in the range of a few minutes to 1 day.
• The majority of volunteer computing power is in GPUs. If possible, provide GPU apps (CUDA, OpenCL, Metal).

Applications and job submission

• BOINC Central supports Docker-based apps. If you already use Docker to run apps on AWS or Linux clusters, you can use BOINC Central with almost no changes.
• BOINC Central provides a simple web-based system for submitting batches of jobs, monitoring their progress, and retrieving their output files.

Getting started

If you're interested in computing with BOINC Central:

• Create a BOINC Central account. You don't have to install BOINC on your own computer, though we encourage you to do so.
• Apply for computing by filling out this form.